Hi roger2,
Use 500V units. That is a standard value. Many you see in equipment might be 50 V or 100V. Although 100 VDC capacitors might be okay in your set, the rating is very marginal. For some reason, capacitors seem to get better as the voltage increases - to a point. I don't know about what happens to the dielectric when the rated voltages exceed 1 KV. I normally only stock 500 V ceramic capacitors. 300 and 500 volt mica capacitors and usually 630 V polystyrene capacitors. I do have an assortment of 50 V polystyrene, but those are for tuners and preamplifiers. Open reel and cassette decks would need at least 150 V in the bias oscillator section.
Would you please point out what exact parts you are using? Here is a link to Murata's capacitors. http://www.murata.com/~/media/webrenewal/support/library/catalog/products/capacitor/mlcc/c02e.ashx
I don't use these, but I feel uncomfortable with what I think you are using.
Normally you might want to use silver mica, polypropylene or similar type parts. I stay away from all ceramic capacitors unless I am working in high frequency circuits, or I can't find the value I need in any other type of capacitor. This is probably my own prejudice, but I can't change that.
-Chris
Use 500V units. That is a standard value. Many you see in equipment might be 50 V or 100V. Although 100 VDC capacitors might be okay in your set, the rating is very marginal. For some reason, capacitors seem to get better as the voltage increases - to a point. I don't know about what happens to the dielectric when the rated voltages exceed 1 KV. I normally only stock 500 V ceramic capacitors. 300 and 500 volt mica capacitors and usually 630 V polystyrene capacitors. I do have an assortment of 50 V polystyrene, but those are for tuners and preamplifiers. Open reel and cassette decks would need at least 150 V in the bias oscillator section.
It certainly does! The same holds true for the output section transistors.
Part size is always a concern, even in older tube equipment where "Orange drop" capacitors do not fit in the same physical location. If you have a component to fit on a circuit board, it must fit into the existing holes - unless you have a clear path to drill holes further away that allows the leads to connect as if a copper trace was run. If the leads are too large, the pads and traces can tear later on. Larger components like capacitors ought to be glued down with silicone adhesive as well. (solder after gluing!) This removes the stress on the copper pads. The body of the replacement part must not crowd adjacent parts, or disturb airflow to transistors or resistors. Heat build up kills everything, including capacitors. So you see, the size of replacement parts is always a concern. I have had equipment in for service where a person intent on improving things has caused enough damage to write the equipment off as "not worth repair". This applies especially if the damage cannot be repaired to a reliable state. So please keep replacement components to a compatible size. Also, never replace carbon film resistors (or carbon composition) in tuners, servos (like in a CD player) or anything else where the change will upset adjustments or in high frequency circuits.
I would use the next higher voltage rating. 100 V is very marginal.
Especially those! Any other capacitors in the same positions elsewhere in the amplifier or preamp circuits should also be replaced.
Would you please point out what exact parts you are using? Here is a link to Murata's capacitors. http://www.murata.com/~/media/webrenewal/support/library/catalog/products/capacitor/mlcc/c02e.ashx
I don't use these, but I feel uncomfortable with what I think you are using.
Normally you might want to use silver mica, polypropylene or similar type parts. I stay away from all ceramic capacitors unless I am working in high frequency circuits, or I can't find the value I need in any other type of capacitor. This is probably my own prejudice, but I can't change that.
-Chris
Hi jooch,
-Chris
Wasn't there an issue with the output impedance from the selector side and into the amplifier circuit? What the tone circuit did for you was to buffer the signal, and most tone circuits have gain. A flat gain stage would have been the correct replacement for your ex-tone stage in that case. Something simple using an NE5532A would be fine.
-Chris
The correct order of the boards is numerical in ascending order.
The original sequence was selector, volume/balance, tone control, power amplifier - M3 numbers 4501A, 4502A/B (B is an annex), 4503 and 4504.
If you want to bypass the tone control the feed has to come from 4502A and you will have to remove your 4503 in and out signal connections from the chain. You could make them switchable if you are fussy about retaining the facility - in that case ensure 4503 is in the correct sequence in the chain.
I noticed there are black earth wires in the circuit diagram which are absent in your block diagram you have depicted earth connections but no color identifier.
Last edited:
Hi mjona,
Let's not get picky about wire colours and earth symbols. While the diagram is drawn using the symbol for earth ground (unfortunate choice), we can pretty much assume that the diagram depicts the common "ground" inside the equipment. They do not intend these points to be a literal earth ground. The lack of a colour isn't important, and most diagrams are drawn without them.
Your observations on the order of the circuit boards should help everyone follow along once jooch has responded to your question.
-Chris
Let's not get picky about wire colours and earth symbols. While the diagram is drawn using the symbol for earth ground (unfortunate choice), we can pretty much assume that the diagram depicts the common "ground" inside the equipment. They do not intend these points to be a literal earth ground. The lack of a colour isn't important, and most diagrams are drawn without them.
Your observations on the order of the circuit boards should help everyone follow along once jooch has responded to your question.
-Chris
Agreed.
Sorting earth routes in amplifiers is something I have done before.
It is not easy to see what someone's problem is without the physical hardware, but I do have something comparable that might help.
The discrete circuits in this Akai model resemble those a Kenwood KA3500 amplifier I have. It is probably older than the Akai - it has a captive mains 3 core cable. Close to the chassis entry the mains earth is soldered to a safety tab (A) a part of the chassis.
There is a similar terminal tab (B) that is part of the chassis near the phono input terminals.
That forms a singular meeting point for all lines referenced to "ground" (U.S.) or " earth" (U.K.). E.g. All the input terminals have the insulated outer case connections linked in series - to tab (B).
The circuit board "ground"/ "earth" connections are linked by black wires - through to "the common ground" as you have described - at tab (B).
Some of the black wires are in series following the signal path through the preamplifier board. The the final link to tab (B) is a single black wire spiral wrapped around a bundle of input signal leads to the back panel.
If the internal wiring is changed by cutting out sections of the circuit continuity of the path to "common ground" or "earth" at tab(B) is as important as the signal path connections.
The Akai circuit diagram shows wires to "common ground" as being black - detail that is lacking in "Jooch's" block diagram. The possibility is connections have made to an assumed ground series connection on another board and my question was to prompt some reflection on the whole wiring issue without being critical.
The other unknown question is where exactly is the " common ground" in the Akai. It may be near the phono section, possibly at the tape DIN socket but that is something for "Jooch" to do a visual inspection to find out.
I have not readed the whole thread...maybe people have instructed you better than i
can do.....but let's go to practice.
Put a scope in your output and a signal generator to a very low level.... read a few milivolts in the output, then reduce bias till you see crossover in your scope..increase till it becomes good and ready to go.
Optimized bias may be achieved using spectrum analiser and distortion analiser plugged into your set up...not having these ones, just avoid crossover signs of unbiased.
Anatech is a sacred monster related practice.... be attention related all he say.
regards,
Carlos
can do.....but let's go to practice.
Put a scope in your output and a signal generator to a very low level.... read a few milivolts in the output, then reduce bias till you see crossover in your scope..increase till it becomes good and ready to go.
Optimized bias may be achieved using spectrum analiser and distortion analiser plugged into your set up...not having these ones, just avoid crossover signs of unbiased.
Anatech is a sacred monster related practice.... be attention related all he say.
regards,
Carlos
Hey Carlos!
How have you been? I am just returning, so it is good to see you.
Hmmmm, a monster eh? I guess I might seem to be to some people out there. But you are the experienced one who has also learned by personal empirical demonstration! 🙂 I find this way of learning tends to make the message stick, wouldn't you agree?
-Chris
How have you been? I am just returning, so it is good to see you.
Hmmmm, a monster eh? I guess I might seem to be to some people out there. But you are the experienced one who has also learned by personal empirical demonstration! 🙂 I find this way of learning tends to make the message stick, wouldn't you agree?
-Chris
Hi mjona,
-Chris
True enough!
Careful, that can lead you to nasty surprises. Not every designer will route the common path the same way. This is an area where you can see major differences between product lines.
You are absolutely correct!
That is both fine and correct. The wording of the comment didn't seem to convey that message.
Where indeed? These are things that can trip up people starting out, although they usually forget to run the ground entirely.
-Chris
I've update the schematic to:
The common ground (BLK) in the modified situation is still located at the original input.
I do not have additional (audible) noise since the modification.
BTW: Nice vibes going on here this totally moved from a 'simple' repair to a full on restoration job.
Thanks everyone for their input and help so far, i really wasn't expecting this to become so big.
@anatech
Perhaps this topic should be renamed to "AKAI AM-2450: Restoration & Tweaks"
- make it more clear where the common ground is located
- added input resistors (R1)
- added loudness function
The common ground (BLK) in the modified situation is still located at the original input.
I do not have additional (audible) noise since the modification.
An externally hosted image should be here but it was not working when we last tested it.
.gif){kind=link}
BTW: Nice vibes going on here this totally moved from a 'simple' repair to a full on restoration job.
Thanks everyone for their input and help so far, i really wasn't expecting this to become so big.
@anatech
Perhaps this topic should be renamed to "AKAI AM-2450: Restoration & Tweaks"
Here is a picture of the current situation, hope this helps a little in getting an overview. The black common ground is barely visible but just under the blue/white wires.
An externally hosted image should be here but it was not working when we last tested it.
Here is the PCB layout of the tone and volume board.
An externally hosted image should be here but it was not working when we last tested it.
Yes, i agree Anatech.
I am fine, thank you.
Not offering circuits as DIY anymore due to non authorized copies being sold everywhere.
Now doing in a selfish way, just for fun and distributing Dx Super A in Brazil.
I am still here, from time to time i come to read something.
I am glad to see you around.
regards,
Carlos
I am fine, thank you.
Not offering circuits as DIY anymore due to non authorized copies being sold everywhere.
Now doing in a selfish way, just for fun and distributing Dx Super A in Brazil.
I am still here, from time to time i come to read something.
I am glad to see you around.
regards,
Carlos
The stereo-mono switch shorts the two inputs together after 4k7 resistors in uninterrupted feed lines. These components are on board M3 -4502A, and the so designated switch and added 4k7 resistors shown in your circuit block are set up differently, and I don't know why these have been included.
Also there appears to be a parallel signal feed from the selector board to the tone control board M3-4503 which should follow in series from M3-4502A volume. That does not look right.
There are a number of options to bypass the tone control section. It would not be a bad idea to fetch the second AM-2450 unit you own from storage so the modified unit wiring can be compared visually with the standard one.
Hi jooch,
You need to build a transistor beta matcher. I have posted the details a couple of times. Maybe it can be found.
-Chris
Well, this thread flows nicely as it is. You might want to begin that thread and refer to this one. Maybe start that using this Akai as an experimental, then take what is learned on this one and do the other as a planned upgrade right from the beginning. By that time you will know better what worked for you and possibly why.
You need to build a transistor beta matcher. I have posted the details a couple of times. Maybe it can be found.
-Chris
Hi jooch,
I found an old post where the matcher was described. Click on these links and build one up. The 1% resistors are critical, I used 0.1% ones. They don't cost that much these days. Since your readings depend entirely on these resistances being equal, they directly affect the quality of your match.
N side : http://www.diyaudio.com/forums/solid-state/136851-citation-12-help-2.html#post1730316
P side : http://www.diyaudio.com/forums/solid-state/136851-citation-12-help-2.html#post1730319
Perf board construction : http://www.diyaudio.com/forums/solid-state/136851-citation-12-help-2.html#post1730334
The design has not changed since then. Your transistor matches may require you match any emitter degeneration resistors if they exist. The matches are close enough to make those resistors matter. That is pretty tight!
-Chris
I found an old post where the matcher was described. Click on these links and build one up. The 1% resistors are critical, I used 0.1% ones. They don't cost that much these days. Since your readings depend entirely on these resistances being equal, they directly affect the quality of your match.
N side : http://www.diyaudio.com/forums/solid-state/136851-citation-12-help-2.html#post1730316
P side : http://www.diyaudio.com/forums/solid-state/136851-citation-12-help-2.html#post1730319
Perf board construction : http://www.diyaudio.com/forums/solid-state/136851-citation-12-help-2.html#post1730334
The design has not changed since then. Your transistor matches may require you match any emitter degeneration resistors if they exist. The matches are close enough to make those resistors matter. That is pretty tight!
-Chris
Sorry, i was sloppy there. Both signal feeds are branched after the 4k7 resistors.
Not seeing the actual hardware makes things difficult but it really wasn't that complicated:
Input > selector > mono/stereo > balance > loudness/volume > tone > amplifier
Now it is:
Input > selector > source [direct] [tone > balance] > loudness/volume > amplifier
The source selector switch is connected to the output of each source so both sources are fed by the same input signal at all times.
Last edited:
There are problems with the second layout.
The Akai amp uses a tone control that was widely used.
Versions of this seen in later generations had a separate negative feedback path which converted the stage from a tone control with gain to a line stage with gain.
I have a Technics amplifier with this feature - where a push button switch would select whichever negative feedback network to change between modes. The principle is similar to what Anatech said in posts 25 and 28.
I am ambivalent about the benefits of "Direct Source" buttons. If you want to learn more about that you would look for amplifiers with that facility on Hi-Fi Engine e.g. Technics SX-VU820. To adapt this for your purpose you would have to rescale the values of the resistors in the line stage alternative negative feedback loop.
The suggestion to upgrade the tone control capacitors makes sense to me.
I hope having studied the layouts of other amplifiers you will see the preamplifier stages in these follow a common pattern like the original order in your Akai unit. This is an example of matched design elements rather than independent modules which can be shuffled into any preferred order - in series or parallel. Why would you want to compromise in that way?
Last edited:
That's an interesting suggestion, would that bypass the bass and treble pots as well?
I am a very keen listener and one thing i noticed in my simple experiments that can make a huge difference in the overall spatial image is balance. Some years ago i was running an amplifier directly from a computers sound card, very detailed stereo image but dangerous when anyone else was using the system 😉
So i got interested in 80's amplifiers of which most had a source direct function. The magic was back and then i got curious if upgrading the volume pot would improve anything. Well it did. I don't know if anything else is changed other than balance in a potentiometer but in my opinion you should really have no more than one, if possible an SMD (dact style) attenuator.
Therefore i am not really worried by the preamp itself but all the cheap pots it uses which ruin the stereo image. I could of course upgrade these with something else but that would be silly money and labor wise.
But i see where you're going impedance wise, my current setup indeed would be wrong. I looked at (and mimicked) other amplifiers of which i have schematics and this is is how those are wired, they are however compromised of 2 gain stages (op amps).
To fix this i would have to branch the signal after the loudness/volume control, which is possible and not that hard to do since all wiring runs trough screw terminals now.
Alright we'll stick to this one, when the amp is repaired and still subject to modifications i might move to a new topic.
Nice suggestion on the transistor matcher as i was already considering this kit from Ebay:
BJT Transistor Curve Tracer Plus AC DC Regulator Power Dual Supply "Double Pack" | eBay
Would the kit be just as useful for audio?
You mean having set the bias a bit too high could cause distortion too?
Anatech was helpful and right about many components, so i agree 😀. I am very glad he pointed out the replacement of the resistors, to my surprise those made a big difference.
- Status
- Not open for further replies.