Another update on the progress:
Modification of the VU lighting
As i mentioned earlier i wanted a more even distributed lighting on the VU meters. To achieve this i started out with a 10mm warm white LED and then put some deflectors made out of aluminum to direct the light to the front.
The 10mm LED unfortunately wasn't bright enough and the 'warm white' just looked awkward. Then i realized rear bicycle bulbs are very common and just the right amount of power (0.6w). I only had to put a resistor on it to compensate for the higher voltage.
I had lighting again on the VU meters, but it was like the original bulb. So i started fiddling with the reflectors again i made and made some progress.
To minimize loss on the desired area's i wrapped the prism in aluminum tape.
I then put in the deflector and was able to get this, still a little uneven but overall much better.
Main power supply caps
I know this was not recommended but i really felt the power caps needed attention, especially considering age and also the bass was lacking or sort of weak. From my earlier experiences i never had troubles 'upgrading' power caps. I got confused when some pointed out the risk of increased capacitance. From 2x10000 to 4x6800, how can that be such a risk? Extra power cabling could turn into an FM transmitter, started feeling dizzy. 🙁
I then checked Rod's website again concerning power supplies and read about the theoretical ripple with 10000uF which should be around 1V lower at load.
So i put some white noise trough the amp and measured the voltage at 3 watts of power. Voltage went down from 44.3 to 41, not cool.
Some more reading on rod's same article i stumbled on this: Major Myth Regarding Capacitance
So what to believe then? Well the caps were still old and not performing within the theoretical limit, so better to just replace them with something fresh. 🙄
Since i didn't want to spend money on something that would fit the old caps and didn't want to alter the looks of the AKAI's innards (with a hacked circuit board) i resorted to crafting again.
Unfortunately plumbing did not fit the holes of the old caps so i had to make a custom wrap for the new caps.
Putting the power back on was a little scary after all the warnings...
but nothing special happened, at least not until i put on some music again that is 😀
Holy crap!
Hellooo bass!
The messy bass from before was crystal clear again, definition and the attack was back.
Also no FM sounds from the additional 2 inches of wire and no fuses blowing, well that's a relief.
I put some white noise trough the amp again and measured the voltages once more. This time a drop of 1.2v at 5 watts of output. Guess Rod's theory was about right.
Modification of the VU lighting
As i mentioned earlier i wanted a more even distributed lighting on the VU meters. To achieve this i started out with a 10mm warm white LED and then put some deflectors made out of aluminum to direct the light to the front.
The 10mm LED unfortunately wasn't bright enough and the 'warm white' just looked awkward. Then i realized rear bicycle bulbs are very common and just the right amount of power (0.6w). I only had to put a resistor on it to compensate for the higher voltage.
I had lighting again on the VU meters, but it was like the original bulb. So i started fiddling with the reflectors again i made and made some progress.
To minimize loss on the desired area's i wrapped the prism in aluminum tape.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
I then put in the deflector and was able to get this, still a little uneven but overall much better.
An externally hosted image should be here but it was not working when we last tested it.
Main power supply caps
I know this was not recommended but i really felt the power caps needed attention, especially considering age and also the bass was lacking or sort of weak. From my earlier experiences i never had troubles 'upgrading' power caps. I got confused when some pointed out the risk of increased capacitance. From 2x10000 to 4x6800, how can that be such a risk? Extra power cabling could turn into an FM transmitter, started feeling dizzy. 🙁
I then checked Rod's website again concerning power supplies and read about the theoretical ripple with 10000uF which should be around 1V lower at load.
So i put some white noise trough the amp and measured the voltage at 3 watts of power. Voltage went down from 44.3 to 41, not cool.
Some more reading on rod's same article i stumbled on this: Major Myth Regarding Capacitance
So what to believe then? Well the caps were still old and not performing within the theoretical limit, so better to just replace them with something fresh. 🙄
Since i didn't want to spend money on something that would fit the old caps and didn't want to alter the looks of the AKAI's innards (with a hacked circuit board) i resorted to crafting again.
Unfortunately plumbing did not fit the holes of the old caps so i had to make a custom wrap for the new caps.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Putting the power back on was a little scary after all the warnings...
but nothing special happened, at least not until i put on some music again that is 😀
Holy crap!
Hellooo bass!
The messy bass from before was crystal clear again, definition and the attack was back.
Also no FM sounds from the additional 2 inches of wire and no fuses blowing, well that's a relief.
I put some white noise trough the amp again and measured the voltages once more. This time a drop of 1.2v at 5 watts of output. Guess Rod's theory was about right.
Hi jooch,
Your displays don't look any worse than the ones using a light bulb. Way to go! Very nicely done.
Extra bass. Sorry, the added bass from an increase in filter capacitance just doesn't happen in real life. The only time more capacitance will change things is peaking just below clipping, and even then the peak draw happens in short bursts.
So, how comes it you have more bass???! Simple. If your old filter capacitors are in fact "tired", a new replacement with the same capacitance would result in the same increase in bass. I have had to prove this to people using both speakers and instruments. No difference except for a fraction of a dB. Audible? Probably not since you have to be close to clipping but just short of it. In practice, attempting to do this by ear almost always results in minor clipping.
Let's examine what happens to the rectifiers and filter caps being increased. This is physics folks, pure and simple. Say you are delivering 5 watts to the load (your speakers). You are delivering a specific amount of power to the load irrespective of how much supply voltage there is. 5 watts is 5 watts from your receiver of a Carver TFM-75. A gross comparison to illustrate the point. So your load receives 5 watts, then you quadruple the capacitance. Okay, so the load still is receiving 5 watts of power. However, since you have noticed (thank you for being attentive), the supply voltages do not sag. That means that there is more voltage across the amplifier, so the wasted power is increased. Now, this is where Rod may be confused a little. The load receives 5 watts of power, call the AB amp 50% efficient, so the total power delivered by your power supply is 10 watts. I'm just trying to put a realistic number on the power so you can figure out output power vs input power to the amplifier. At just below clipping, AB amps are at their highest efficiency. It's worse as the power levels decline. Anyway, 10 watts to the amplifier circuits must be balanced by exactly 10 watts back into the supply capacitors less a tiny fraction in case conduction occurs during a charging pulse. Since the voltage across the capacitors is higher, the amount of time that charging current can flow is reduced. Also, as current through a diode increases, so does the voltage drop. That decreases the power charge time even more.
Alright, so we do have a constant amount of power needing to be replaced. We have a shorter amount of charge time in which to do this, which forces the peak currents to be *much* higher. This causes increased voltage drop across the diodes, and this effect can make the capacitance look much, much higher to the transformer and rectifiers. You decreased your voltage drop from 3.3 VDC to 1.2 VDC. Compare that to the increased voltage drop, depending on rectifier could be 0.6 VDC (0.6 V to 1.2 V or even higher). The simulator does not correct for this normally, and so it's effect is not figured in. The net result of that nasty surprise is still a shorter conduction angle (compared to 360 °). The peak currents will be much higher, and an FFT of those peaks will be increasingly higher fundamental frequency.
Measuring these effects will not happen with the average DMM because their frequency response isn't nearly high enough. This also requires a current pickup, typically used with high performance oscilloscopes. Using the correct equipment will show you that the extreme current peaks are much higher than you would anticipate. These cause something called an IR loss in the diodes and transformer windings. This does result in power transformers that run hotter than they normally would, but only in a sustained high power state. A state that you might see in a peeler bar, or at home on New Year's Eve. Normal running will warm it up slightly, but not enough to notice. This will only bite you when you really need things to run cooler.
Now, the slightly higher supply voltages, what do they really buy you? Almost nothing. Go ahead and figure out what your peak power is under nearly clipping conditions with the designed in capacitance, and the increased capacitance. Under those conditions, expect maybe a single dB. You cannot hear this, you can not sense this, it makes exactly zero difference to you with your ears.
All I am saying hear is that you have to put things into perspective. What you have been told here is the truth in all it's glory.
With sensitive equipment you certainly will see an increase in radiated noise with higher capacitance. Just ask the poor folks who failed their EMI/RFI acceptance tests. Will you hear it? Maybe yes in an AM radio, possibly it might affect an MPX demodulator (maybe). However, you probably wouldn't notice, and if you did you wouldn't be able to figure it out. Nothing is doom and gloom, but with increased capacitance and extended leads you absolutely are radiating more energy. This will become more apparent when the system is cranked. Keep in mind too, the RMS to peak power ratio often ranges from 15 to 20. That means your average power on the meter does not show you the peaks that might be 15 to 20 X that indicated value. A consumer peak reading meter generally reads lower than the true peaks. For peak indication you need a light/LED driven with a pulse stretcher so your eye can respond to the otherwise brief flash that would occur. At no time were we worried about blowing fuses or FM being affected. The noise generated is mostly AM in nature which an FM receiver is not sensitive to.
I hope that helps explain the differences for you in a way that can be reasoned through. Rod Elliot has done you a disservice by claiming these effects do not exist, because they are real. I have had to deal with the increased capacitance craze that swept through Ontario for years. I just got a call from someone who had their preamp "modded" with higher capacitance filter caps. Arced the rectifier. Excellent work by the internet "mod squad". Now I have to clean up another mess that Rod claims isn't a true concern. The owner of the preamp is certainly concerned. Maybe Rod should walk in the shoes of experienced service people who have also had the benefit of higher education on the topic. The damage done to this man's preamp was avoidable.
Rod's theory is unfortunately incorrect, proved empirically by a third party.
-Chris
Your displays don't look any worse than the ones using a light bulb. Way to go! Very nicely done.
Extra bass. Sorry, the added bass from an increase in filter capacitance just doesn't happen in real life. The only time more capacitance will change things is peaking just below clipping, and even then the peak draw happens in short bursts.
So, how comes it you have more bass???! Simple. If your old filter capacitors are in fact "tired", a new replacement with the same capacitance would result in the same increase in bass. I have had to prove this to people using both speakers and instruments. No difference except for a fraction of a dB. Audible? Probably not since you have to be close to clipping but just short of it. In practice, attempting to do this by ear almost always results in minor clipping.
Let's examine what happens to the rectifiers and filter caps being increased. This is physics folks, pure and simple. Say you are delivering 5 watts to the load (your speakers). You are delivering a specific amount of power to the load irrespective of how much supply voltage there is. 5 watts is 5 watts from your receiver of a Carver TFM-75. A gross comparison to illustrate the point. So your load receives 5 watts, then you quadruple the capacitance. Okay, so the load still is receiving 5 watts of power. However, since you have noticed (thank you for being attentive), the supply voltages do not sag. That means that there is more voltage across the amplifier, so the wasted power is increased. Now, this is where Rod may be confused a little. The load receives 5 watts of power, call the AB amp 50% efficient, so the total power delivered by your power supply is 10 watts. I'm just trying to put a realistic number on the power so you can figure out output power vs input power to the amplifier. At just below clipping, AB amps are at their highest efficiency. It's worse as the power levels decline. Anyway, 10 watts to the amplifier circuits must be balanced by exactly 10 watts back into the supply capacitors less a tiny fraction in case conduction occurs during a charging pulse. Since the voltage across the capacitors is higher, the amount of time that charging current can flow is reduced. Also, as current through a diode increases, so does the voltage drop. That decreases the power charge time even more.
Alright, so we do have a constant amount of power needing to be replaced. We have a shorter amount of charge time in which to do this, which forces the peak currents to be *much* higher. This causes increased voltage drop across the diodes, and this effect can make the capacitance look much, much higher to the transformer and rectifiers. You decreased your voltage drop from 3.3 VDC to 1.2 VDC. Compare that to the increased voltage drop, depending on rectifier could be 0.6 VDC (0.6 V to 1.2 V or even higher). The simulator does not correct for this normally, and so it's effect is not figured in. The net result of that nasty surprise is still a shorter conduction angle (compared to 360 °). The peak currents will be much higher, and an FFT of those peaks will be increasingly higher fundamental frequency.
Measuring these effects will not happen with the average DMM because their frequency response isn't nearly high enough. This also requires a current pickup, typically used with high performance oscilloscopes. Using the correct equipment will show you that the extreme current peaks are much higher than you would anticipate. These cause something called an IR loss in the diodes and transformer windings. This does result in power transformers that run hotter than they normally would, but only in a sustained high power state. A state that you might see in a peeler bar, or at home on New Year's Eve. Normal running will warm it up slightly, but not enough to notice. This will only bite you when you really need things to run cooler.
Now, the slightly higher supply voltages, what do they really buy you? Almost nothing. Go ahead and figure out what your peak power is under nearly clipping conditions with the designed in capacitance, and the increased capacitance. Under those conditions, expect maybe a single dB. You cannot hear this, you can not sense this, it makes exactly zero difference to you with your ears.
All I am saying hear is that you have to put things into perspective. What you have been told here is the truth in all it's glory.
With sensitive equipment you certainly will see an increase in radiated noise with higher capacitance. Just ask the poor folks who failed their EMI/RFI acceptance tests. Will you hear it? Maybe yes in an AM radio, possibly it might affect an MPX demodulator (maybe). However, you probably wouldn't notice, and if you did you wouldn't be able to figure it out. Nothing is doom and gloom, but with increased capacitance and extended leads you absolutely are radiating more energy. This will become more apparent when the system is cranked. Keep in mind too, the RMS to peak power ratio often ranges from 15 to 20. That means your average power on the meter does not show you the peaks that might be 15 to 20 X that indicated value. A consumer peak reading meter generally reads lower than the true peaks. For peak indication you need a light/LED driven with a pulse stretcher so your eye can respond to the otherwise brief flash that would occur. At no time were we worried about blowing fuses or FM being affected. The noise generated is mostly AM in nature which an FM receiver is not sensitive to.
I hope that helps explain the differences for you in a way that can be reasoned through. Rod Elliot has done you a disservice by claiming these effects do not exist, because they are real. I have had to deal with the increased capacitance craze that swept through Ontario for years. I just got a call from someone who had their preamp "modded" with higher capacitance filter caps. Arced the rectifier. Excellent work by the internet "mod squad". Now I have to clean up another mess that Rod claims isn't a true concern. The owner of the preamp is certainly concerned. Maybe Rod should walk in the shoes of experienced service people who have also had the benefit of higher education on the topic. The damage done to this man's preamp was avoidable.
Rod's theory is unfortunately incorrect, proved empirically by a third party.
-Chris
Definitely not more bass, but better bass. Before it was just muddy or blurry, now it is tight and focused again in such a way that you notice it immediately (at just normal listening levels).
I see your point
I think this is also part of the truth, most people will not notice the actual effects on the power supply because not many will push their equipment. I rarely listen to anything over 4 watts of power, getting hearing damage once was enough for me 😉
Well yes, and i never doubted there wouldn't be any truth in that. It is however an inconvenient one if you are trying to keep the forward momentum going on concerning the restoration.
I know in the end Rod is just another enthusiast, but he does have some practical projects to learn from and presents the information in an uncomplicated way anyone can pick up. Many tend to forget how much work it is to write material and maintain a website by yourself.
A little more than a decade ago Hi-Fi/TV shops all had someone who did repairs. The good ones asked prices that seemed high to some and so much equipment ended up with the dodgy shops. The repairs i have seen made by these guys are truly horrific.
I bought a beautiful National Panasonic SU-3100B for a few euro's some time ago which was at some time at a 'repair' shop. The 'repair' guy couldn't figure the circuit out and just ripped everything from the amplifier board (including some of the board itself) except the power supply parts and then just hacked in (plumbers way) an amp board from another (likely destroyed) amp.
The sorts of upgrades you see today are just another form of abuse. My AKAI was upgraded (damaged) with an LED on the volume knob (hole drilled) so you can see the volume level, dear god...
There's no way to stop people from doing stupid things, but most happen because people don't take the necessary time to do something right (maybe reconsider) or admit their inadequate skill level to do so.
I now know the negative affects of increased capacitance which i will definitely consider in other jobs, but i will also remember how much this one has gained in musical value from all the new parts.
Hi jooch,
Well, I am hoping that what is written here is complete and correct, because several other people will read it also. I don.t worry so much about what you are doing, because the work you did was neat and safely done. Just try to check the temperatures on your filter caps in case they can't get rid of the heat that is generated in normal operation.
Most people read information off the internet in various places in order to answer a question. They assume everything is correct and stop reading when they figure their question has a satisfactory answer. Most will not check around any further and simply launch into a repair/upgrade without any further consideration to what is actually going on.
Like I said, the larger capacitors will not kill your receiver, but your power switch might be helped if you install an M.O.V across the transformer primary, directly across the primary (electrically) behind the fuse and power switch. What this does is greatly extend the life of your power switch because it prevents the contacts from arcing on turn-off. For turn on you will see either a surge M.O.V. (different beast) in series with the primary, or a power resistor that is shorted out after a second or two. This reduces the inrush current. Your lights would not flicker anymore on turn on. Just stick the coil of the relay across the DC main power supply. When the supplies come up enough, the coil actuates and shorts out the resistor. There you go, effective inrush current protection.
Many high powered amplifiers and receivers have this built in. You turn them on <click>, a second <click> is heard after a second or two, then the third and final <click> as the speaker relay closes. The second click was the surge resistor being shorted out.
Yes, you really improved the performance. Now, get rid of the crappy capacitors. All those ceramic capacitors (without a black stripe on top) and tired electrolytic capacitors. This will bring your set closer to an amplifier you will really enjoy listening to.
-Chris
Well, I am hoping that what is written here is complete and correct, because several other people will read it also. I don.t worry so much about what you are doing, because the work you did was neat and safely done. Just try to check the temperatures on your filter caps in case they can't get rid of the heat that is generated in normal operation.
Most people read information off the internet in various places in order to answer a question. They assume everything is correct and stop reading when they figure their question has a satisfactory answer. Most will not check around any further and simply launch into a repair/upgrade without any further consideration to what is actually going on.
Like I said, the larger capacitors will not kill your receiver, but your power switch might be helped if you install an M.O.V across the transformer primary, directly across the primary (electrically) behind the fuse and power switch. What this does is greatly extend the life of your power switch because it prevents the contacts from arcing on turn-off. For turn on you will see either a surge M.O.V. (different beast) in series with the primary, or a power resistor that is shorted out after a second or two. This reduces the inrush current. Your lights would not flicker anymore on turn on. Just stick the coil of the relay across the DC main power supply. When the supplies come up enough, the coil actuates and shorts out the resistor. There you go, effective inrush current protection.
Many high powered amplifiers and receivers have this built in. You turn them on <click>, a second <click> is heard after a second or two, then the third and final <click> as the speaker relay closes. The second click was the surge resistor being shorted out.
Yes, you really improved the performance. Now, get rid of the crappy capacitors. All those ceramic capacitors (without a black stripe on top) and tired electrolytic capacitors. This will bring your set closer to an amplifier you will really enjoy listening to.
-Chris
@ anatech
why the suggestion to replace ceramic caps? typical advice i have seen in forums says to leave them alone
why the suggestion to replace ceramic caps? typical advice i have seen in forums says to leave them alone
As far as i can guess they don't age much, but better tolerances and materials are available nowadays. Old amplifiers like mine are usually not fitted with anything better than 10% tolerance.
Been busy again and have now completed the rewiring to bypass the preamp (i.e. source direct), only the volume control is used in source direct mode.
I was immediately struck by the difference in gain of the amplifiers, only with the volume fully opened do the channels match up. Darn, another component that needs replacing.
Well i ordered a high-end ALPS to replace the standard one, but the shaft is 5mm shorter so this could well become another project 🙄
The mod to a source direct mode really paid off and is able to get the last details (sound wise) out of the amp. The only problem i currently have is when switching modes i have a little pop, can i reduce this with a ceramic cap?
Input Stage
Another thing that still needs attention are the transistorpacks (2SA979) in the input stage. I have ordered some 2SA970's to replace them, but these are not a straight fit and also expensive.
The BC556B is very common here and i can get 10 for less than 2 euro's. With that amount it's possible to match them, also the pin configuration makes them a straight fit to the original holes.
The only problem is the VCE, it needs to be lowered a little as far as i can tell from the schematic. R4, R5 and R6 make up a voltage divider that determine the VCE, correct?
If so can i reduce R5 to lower the VCE?
I was immediately struck by the difference in gain of the amplifiers, only with the volume fully opened do the channels match up. Darn, another component that needs replacing.
Well i ordered a high-end ALPS to replace the standard one, but the shaft is 5mm shorter so this could well become another project 🙄
The mod to a source direct mode really paid off and is able to get the last details (sound wise) out of the amp. The only problem i currently have is when switching modes i have a little pop, can i reduce this with a ceramic cap?
Input Stage
Another thing that still needs attention are the transistorpacks (2SA979) in the input stage. I have ordered some 2SA970's to replace them, but these are not a straight fit and also expensive.
The BC556B is very common here and i can get 10 for less than 2 euro's. With that amount it's possible to match them, also the pin configuration makes them a straight fit to the original holes.
The only problem is the VCE, it needs to be lowered a little as far as i can tell from the schematic. R4, R5 and R6 make up a voltage divider that determine the VCE, correct?
If so can i reduce R5 to lower the VCE?
First you should verify by measurement that it actually is the pot - it might also be a dried-out coupling cap or bad solder joint after all.
Is the original a sealed pot?
It is actually possible to restore these things (take apart, clean, seal, put back together), you should merely feel comfortable drilling out smallish rivets as that often needs to be done - not being afraid of things small and mechanical definitely helps. Some experience with contact cleaning and sealing is also recommended.
I'd definitely prefer doing that over a botch job with a new pot that is not an ideal fit mechanically. If it can't be done any other way, well, so be it, but it's better avoided if one can help it.
No.
Please show on the schematic exactly what your mod does, and we may be able to tell which cap's leakage current or whatnot is causing the issue.
Do you have any issues with excessive DC offset then? Otherwise I would strongly advise not to attempt to fix what isn't broken in the first place. Yes, these sometimes fail, but that doesn't mean that yours are begging for replacement right now. (On the right-hand side of the bathtub curve, everything is possible.) That's typical distortion of perception - you only get to read about problems on the interwebs, but you can never tell how many people don't have any issues.
It's a bit like harddrives - even ones known as being of only average reliability can still last a long time. I recently replaced an old Maxtor that had had a file gone unreadable, after about 8 long years of faithful service. A second one is still churning along, noisy as ever and a bit slow by modern standards.
No, no, and no.
Attempting to fix things that aren't broken is particularly recommended against when you don't have a clue of what you're doing.
While a transistor type with Vbr,CEO > (V+ - V-) is a safe bet, it is certainly possible to use one with little more than half that for the input LTP, and in fact this was regularly done up to the 1970s as there simply weren't any suitable higher-voltage transistors available at the time. It takes a bit of care to ensure that the transistors cannot get fried during a fault condition (like when the output goes right up to positive rail voltage), but in this case C5 should do a decent job as a protector. BC556s are rated at 65 V Vbr,CEO, rails are +/-46 V nominal (with LTP positive supply being +35.5 V nominal), so that should work.
You may want to do some reading on differential amplifiers there, in whatever the standard textbooks on transistor circuitry may be where you live. But for illustration, just leave out the right-hand transistor in the dual package and look at input - R5 - TR1 (left) - R4:
TR1 (left) is what kind of device (active/passive)?
What is TR1 (left) base voltage?
What is TR1 (left) emitter voltage, and why?
Where do you expect TR1 (left) collector voltage?
Last edited:
Nope, completely open. Checked the resistance and its badly out of balance, already sprayed some cleaning solution but did not do much.
I agree, but apart from the shaft length the new one will totally fit the original space. I don't have to move any parts or hack any holes 😉
The shaft length can be fixed by taking apart a cheap knob and then add that to the original volume knob with epoxy resin and some fiberglass reinforcement. I have done this many times and works great. Because of the reinforcement the knob will last even if pressure is applied.
I have to draw it because it was quite some work to figure it all out, then again it might also be interesting to others. So i will try to finish the schematic this weekend.
Good question, short answer is no. I do have a problem that when the amp is cold, i have to wait an hour or so before everything is in the 'on' state and starts to conduct signals. At first i suspected the preamp but this can also be ruled out because of the source direct mod.
One 'solution' is to increase the signal strength briefly and it will conduct but only for a while and then the signal disappears again. This problem was always an issue even when the previous owner still had it, however if warmed up it will work fine.
I have already replaced every cap and resistor that conducts the signal, so that is ruled out then i guess.
I indeed still know too little of such circuits to properly choose a replacement, that's why i like to be on the safe side and share my thoughts on these ideas so i (and others) hopefully can learn to do so.
So how does one figure out properly what voltage TR1 sees/gets?
- Passive
- 0.1V
- 0.7V, because of R5 to bias TR1?
- on the emitter
Does this mean that TR1 only 'sees' 35.5V?
The circuit diagram provides a guide to the voltage readings.
You mention in your initial post that you have two of these model of receiver - if the first is working then you can take the cover off and note the readings from that for comparative purposes - to be methodical you would compare these to each other and the circuit diagram and note on paper.
The gain of the power amplifier section is R7+R3/R3 or about 14 times. This depends on the integrity of capacitor C5 and the end of R3 that should connect to earth being intact.
If this line is open there is no divider action and R3 equates to zero and the power amplifier has a voltage gain of 1, which would possibly explain some stability issues that were subject to some other discussion in this thread.
Check the line D which leads to the tone control board is still connected as indicated in the circuit diagram - with zero intermediate resistance.
If not you might look for hairline cracks in the board or from the notes on voltage readings press suspect parts or the board with your finger - repair or re- solder a "cold joint" as appropriate. If temperature related you might try directing some warm breath or gentle heat from a hair dryer through a mailing tube.
The other capacitor which affects voltage amplification is C7 which is part of a "bootstrap" circuit.
By the way, the line/tone stage presents a high impedance to the volume, balance and loudness section. It also has a low output impedance in relation to that of the power amplifier input.
Last edited:
That is a good idea, but out of my reach right now. I was redecorating the house and moved all audio projects to a storage unit in boxes. This one happened to be left behind and that is why i am fixing it right now, i want to finish all projects one by one which helps me to remain focused.
I can write down the voltages on this one though. Would that help?
Thank you for that explanation, learned something 😉 The earth connection is intact and D line still leads to the tone control board without any measurable resistance.
I replaced and resoldered all resistors/electrolytes and while i was at it i also reworked the solder points of TR1. I do have crackling noise when the amp is still cold, from what i know that usually indicates a bad transistor.
I know. That's the reason why i wanted a second preamp without tone control. I was thinking of using a simple NE5532 opamp circuit with a simple zener supply from the +-45V rails or using voltage divider on the 35V rails. Is there any difference performance wise?
Last edited:
Hi jooch, roger2,
Ceramic capacitors that are not NP0 or C0G tend to distort the signal appearing across it. These low value components can really cause a decrease of audio quality. Therefore, it is in fact wise to replace these parts anywhere in the signal path. The ones in the feedback network, or in high impedance circuits (like CDOM) in an amplifier will cause the most distortion. Replace with the same value part. Once you have tried this once, you will do this every time in the future as a normal step.
I can't comment on what and where you might read information on the internet, but a lot of information is mixed up, confused or may be a part in another person's get rich scheme. Non-technical do-it-yourself folks are more often at risk for believing things that are not factual. When you read about flowery descriptions of sonic changes, its time to carefully review what is being claimed. Ask a person who really does know what they are doing if in doubt. I read suggestions every day that are grossly incorrect. Some modifications are even likely to cause damage to equipment.
Input diff pairs should be checked for damage after a failure in an amplifier channel. They are often damaged and might show changes in beta, or increased noise if they break down reverse E-B. The dual transistors have historically not been the best match, and the tightness of the transistor match in a differential pair (long tailed pair) has a direct effect on both distortion and DC offset. Change the darned things. 2SA970 is rated for a 120 VDC emitter - collector breakdown voltage and 100 mA collector current. The 2SA979 is rated for 100 VDC and 50 mA collector current. Therefore the 2SA970 is a very reasonable transistor to use to replace the 2SA979. Two emitter leads go into the center hole. The next two are collectors and the outside pair should be the base leads. Make the cases touch, and maybe use heat shrink tubing to tie the cases together to match their temperature. You will either have to buy a matched pair from someone, or match them yourself. You need the low noise attribute of the 2SA970 or equivalent. This is a very easy fix to perform.
As you have proved to yourself (and I'll believe your results), go ahead and replace the volume control. Those Alps or Nobel controls are much better for channel tracking.
-Chris
Ceramic capacitors that are not NP0 or C0G tend to distort the signal appearing across it. These low value components can really cause a decrease of audio quality. Therefore, it is in fact wise to replace these parts anywhere in the signal path. The ones in the feedback network, or in high impedance circuits (like CDOM) in an amplifier will cause the most distortion. Replace with the same value part. Once you have tried this once, you will do this every time in the future as a normal step.
I can't comment on what and where you might read information on the internet, but a lot of information is mixed up, confused or may be a part in another person's get rich scheme. Non-technical do-it-yourself folks are more often at risk for believing things that are not factual. When you read about flowery descriptions of sonic changes, its time to carefully review what is being claimed. Ask a person who really does know what they are doing if in doubt. I read suggestions every day that are grossly incorrect. Some modifications are even likely to cause damage to equipment.
Input diff pairs should be checked for damage after a failure in an amplifier channel. They are often damaged and might show changes in beta, or increased noise if they break down reverse E-B. The dual transistors have historically not been the best match, and the tightness of the transistor match in a differential pair (long tailed pair) has a direct effect on both distortion and DC offset. Change the darned things. 2SA970 is rated for a 120 VDC emitter - collector breakdown voltage and 100 mA collector current. The 2SA979 is rated for 100 VDC and 50 mA collector current. Therefore the 2SA970 is a very reasonable transistor to use to replace the 2SA979. Two emitter leads go into the center hole. The next two are collectors and the outside pair should be the base leads. Make the cases touch, and maybe use heat shrink tubing to tie the cases together to match their temperature. You will either have to buy a matched pair from someone, or match them yourself. You need the low noise attribute of the 2SA970 or equivalent. This is a very easy fix to perform.
As you have proved to yourself (and I'll believe your results), go ahead and replace the volume control. Those Alps or Nobel controls are much better for channel tracking.
-Chris
1.You could measure the voltages and write them down. For that I suggest making a photocopy of the circuit and with a colored pen, write the voltages found alongside the designated values.
2.When you can get your other receiver out of storage you can repeat the exercise using a different colored pen.
3.While you have measured the power amp earth through the D line and it might be assumed that the tone control board connects to earth, can you confirm that you followed the earth point there to the chassis. There should be zero volts between these two points. E.g. If the tone control board has been bypassed with the leads disconnected, is the earth connection through the shielding braids?
4.If a transistor is working correctly it will have a base to emitter voltage close to 0.6 volts - a useful test to identify problem areas in circuits. I say that since it is not unknown for a transistor to be giving a wrong reading and replacing the part changes nothing.
5.The crackling noise indicates there is some amplification however the problem could be external to the amplification block such as a floating or intermittent connection.
6.When there is an intangible problem which remains unsolved there is an old English expression " I cannot put my finger on the problem".
7.The form of the Akai circuit is an operational amplifier with two inputs both of which are tangible ( the bases of dual transistor TR1). Tapping a finger on the top capacitors connected to either of these points should produce some noise response which in the input signal line should be traceable at the same level back to an unplugged line input with volume control at max.
Steps as in 3. and 7. may avoid the more laborious processes if you do these first.
It is your decision about replacing the tone control board with an NE5532. I would leave that until you get all your repair work done and do some subjective comparisons between your unmodified amplifier and the subject one.
Last edited:
Hi mjona,
A defective component further along in the circuit can easily (and often) cause an input and / or Vas transistor to be cut off. So you might even read a reverse bias, or even a 0 V value that can lead you to make an incorrect assumption. Circuits with feedback don't often lend themselves to this test. Feedback causes a fault to affect components before the actual fault.
-Chris
This is probably not what you want to do. Most circuits will be upset by probes placed on sensitive areas like this. The possibility of a probe slipping and causing damage is also extremely high. These tests are not usually worthwhile unless you have good reason to test this. Also, these type of tests are best done from the foil side of the PCB if they must be done.
A defective component further along in the circuit can easily (and often) cause an input and / or Vas transistor to be cut off. So you might even read a reverse bias, or even a 0 V value that can lead you to make an incorrect assumption. Circuits with feedback don't often lend themselves to this test. Feedback causes a fault to affect components before the actual fault.
Normally the chassis common is not directly earthed. Certainly the system common point is rarely directly coupled to earth. A connection to real earth may be caused by a connection to a cable connection if you have a VCR, TV or FM antenna connection to a cable company for reception of TV or FM signals. This connection is often responsible for horrible lightening damage as the average audio system does not have a connection to earth ground. If you do have such a connection to a cable service, I implore you to find and use "ground break" devices on the cable. This is simply a 1:1 transformer used to isolate the cable ground from your equipment. If you also have a hum in your system, this device will often eliminate that problem as well if it is caused by the ground connection of a cable TV service.
-Chris
Thank you anatech. Lots of interesting points in your posts...
The portion of your post quoted above is an idea which I have been considering for a while (suggested by a friend on another forum) and have been attempting, unsuccessfully, to research further. People seem to love talking about electrolytic and film caps, but have little interest in ceramics.
-is CDOM the same as Miller cap in VAS stage? In this amp, which would be CDOM?
-what voltage spec would you suggest for ceramics for these positions (NFB and CDOM)? any specific brand/type?
-the OP's Akai was made probably in the late 70's or 1980. Are you saying that ceramics used in Japanese amps of that era were not NP0? Or, if they were NP0 at that time, would today's NP0 caps be better?
-would there be any risk of introducing oscillation or other instability by employing this component upgrade?
In post #49 "Jooch" took some measurements around TR1 and worked out the voltage drop across the relevant transistor junctions. He expressed this as what the transistor "sees or gets". Sight is the primary means to steer clear of danger and it appeared that "Jooch" was cognizant of the contact risks with a probe and took his measurements from some point in series with the transistor leads.
This is the method I would use - not by measuring directly between a pair of leads with a probe on each.
With regard to sensitivity of circuits to probes your comments may deter readers from taking any measurements - which begs the question as to why Akai bothered to include indicative voltages on the circuit diagram.
My comments about the earth arrangements are based on the version of the equipment marketed in Europe as " Jooch" resides in the Netherlands. The versions sold in the U.K. and Europe both have three core mains lead connections - phase (brown), neutral (blue), and earth (Green and Yellow). The latter is connected to a copper stake in the ground outside the property.
We follow this convention in New Zealand. Unless equipment meets double insulated standards here as in the U.K. and Europe the case must be connected to the earth terminal on the mains input socket.
Last edited:
Hi roger2,
The ceramic capacitors in audio equipment is generally not C0G / NP0 as these types are about as much as the best capacitor types. The characteristic varies for the generic ceramic capacitors. Note that European and North American manufacture was every bit is bad. Normally you should use a silver mica, polycarbonate or other similar "low K" dielectric. Polystyrene capacitors are one of my favorite types to use, but they are sensitive to heat. Mica is a good high temperature capacitor. The good NP0 / C0G ceramic capacitors should work as well.
The C0G /NP0 ceramic capacitors have existed for a very long time. They are often used in RF work where stability is important. They were also made with specific temperature coefficients to offset the drift of other parts (like a tuning capacitor). These are certainly not a new type of capacitor.
Solid tantalum capacitors are similar to the bad ceramic capacitors. Using an electrolytic capacitor is a step up over the solid tantalum. Wet slug tantalum capacitors are one of the best types to use for larger capacitance values. They are very expensive.
-Chris
Yes, that is exactly what I am talking about. I would consider CE5 and CE9 as critical parts.
The ceramic capacitors in audio equipment is generally not C0G / NP0 as these types are about as much as the best capacitor types. The characteristic varies for the generic ceramic capacitors. Note that European and North American manufacture was every bit is bad. Normally you should use a silver mica, polycarbonate or other similar "low K" dielectric. Polystyrene capacitors are one of my favorite types to use, but they are sensitive to heat. Mica is a good high temperature capacitor. The good NP0 / C0G ceramic capacitors should work as well.
The C0G /NP0 ceramic capacitors have existed for a very long time. They are often used in RF work where stability is important. They were also made with specific temperature coefficients to offset the drift of other parts (like a tuning capacitor). These are certainly not a new type of capacitor.
Solid tantalum capacitors are similar to the bad ceramic capacitors. Using an electrolytic capacitor is a step up over the solid tantalum. Wet slug tantalum capacitors are one of the best types to use for larger capacitance values. They are very expensive.
No brand. The type and quality is far more important than a brand name. Just pick a good industrial brand, not an audiophile approved brand. For voltage ratings, use the highest voltage (within reason) that you can neatly mount in the physical space you have. Installing a larger component such that it doesn't fit well introduces a whole bunch of other problems. You can also damage the PC board traces. Don't do this. One thing that is for sure, you must install a replacement with at least the original part's voltage breakdown rating. If you must use a solid tantalum due to space constraints, at least double the original voltage rating. I try to use the highest voltage rating I can get for these capacitors.
I answered the first part of your question above. The second? All components have improved due to natural improvements being made all the time for any manufacturing. You can get both good and bad quality from almost any country in the world.
No, you should be absolutely safe as long as you are replacing the same part value with good quality parts.
-Chris
Hi mjona,
What are you measuring? Without using an oscilloscope, you have no clue as to whether the voltages are constant, or if there is an asymmetric waveform causing your meter to read whatever it is. Any noise from the meter circuit will be amplified. Just the capacitance of your finger or probe can make the circuit unstable and send it into oscillation. That also invalidates the numbers you see on the meter screen.
Don't forget that service manuals are not written for people who are not completely and properly trained. Manuals fail to mention huge amounts of information that any untrained person will need. Just because there are numbers on a schematic, that does not mean that you can plunk a meter on those test points. Fat, dumb and happy. Most of those measurements are wrong by a small amount. You do not know if they were measured with a meter, a meter with a compensated probe, an oscilloscope. They may even be calculated values. You can not assume that you are safe to go ahead and stick meter probes between those points because those values might be wrong, and may have been arrived at by any of a number of methods. You just don't know.
It is very important to note that you have to pick a suitable connection to use as a reference for a meter or oscilloscope connection. Earth ground is possibly the worst point you could use as your reference. Too many unknown currents will be imposed on that connection between the circuit common and earth.
-Chris
True, but this is still not a safe thing to do. I will not measure these points without a very good reason, and for the same reasons I posted about.
Well, I am not going to assume anything about another persons capabilities. I have seen my share of blown circuits caused by probing these areas - by experts in the field! This is a dangerous practice no matter who you are. Therefore, do not do this unless you must have the measurement. Remember that in a circuit with feedback, problems upstream can throw off bias conditions right at the beginning of a circuit.
What are you measuring? Without using an oscilloscope, you have no clue as to whether the voltages are constant, or if there is an asymmetric waveform causing your meter to read whatever it is. Any noise from the meter circuit will be amplified. Just the capacitance of your finger or probe can make the circuit unstable and send it into oscillation. That also invalidates the numbers you see on the meter screen.
The voltages are for the use of skilled technicians and engineers who understand what effect their equipment may have on a circuit. Most people I know who must measure these points solder on test points to the foil. I do that with things I am working on, as do most people in the industry.
Don't forget that service manuals are not written for people who are not completely and properly trained. Manuals fail to mention huge amounts of information that any untrained person will need. Just because there are numbers on a schematic, that does not mean that you can plunk a meter on those test points. Fat, dumb and happy. Most of those measurements are wrong by a small amount. You do not know if they were measured with a meter, a meter with a compensated probe, an oscilloscope. They may even be calculated values. You can not assume that you are safe to go ahead and stick meter probes between those points because those values might be wrong, and may have been arrived at by any of a number of methods. You just don't know.
Right you are. However, I doubt they are also connected to the internal circuit ground. Here in North America, direct earth connections create all kinds of trouble. Measure between the common point of the main filter capacitors to the chassis, then also to the ground connection on the AC power plug. Then you will know for sure.
Exactly. But then, is the case connected to circuit common with a hard short? Probably not. Like I said, check.
It is very important to note that you have to pick a suitable connection to use as a reference for a meter or oscilloscope connection. Earth ground is possibly the worst point you could use as your reference. Too many unknown currents will be imposed on that connection between the circuit common and earth.
-Chris
Thank you anatech. This is great info!
Re voltage: There is no voltage spec for the ceramic caps anywhere for this amp. Not on the schematic, the parts list, or the caps themselves. I do not understand the circuits well enough to make the call on this.
The rails are ± 47vdc. Does that mean the bypass cap across the large value NFB resistor could possibly see up to 94v?
Part size is not a concern, there is adequate room. I already have 100v MLCC replacements (Murata) for the NFB bypass caps, and they are physically very small. Should I go higher than 100v or would this be adequate for NFB and the other positions?
Lastly, if I am going this far, what about Ce7 and/or Ce11? Is there anything to gain by changing them as well?
Re voltage: There is no voltage spec for the ceramic caps anywhere for this amp. Not on the schematic, the parts list, or the caps themselves. I do not understand the circuits well enough to make the call on this.
The rails are ± 47vdc. Does that mean the bypass cap across the large value NFB resistor could possibly see up to 94v?
Part size is not a concern, there is adequate room. I already have 100v MLCC replacements (Murata) for the NFB bypass caps, and they are physically very small. Should I go higher than 100v or would this be adequate for NFB and the other positions?
Lastly, if I am going this far, what about Ce7 and/or Ce11? Is there anything to gain by changing them as well?
Here's the schematic for the source direct mod.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- Not open for further replies.