I have a couple of Onkyo TX-NR906 AV recievers that I decided to try and improve after testing a few newer amps without getting the performance I was looking for.
I've done the D/A section and low level analog path as well as the power end of the amp section and have started to look at the driver amp section to see if there is potential. The mods are generally just get rid of as many caps is the signalpath as you can, if you need them add a film cap and beef up powersupply and decouple with film caps. Finally make sure the grounding is solid. In the picture you see the driver amp section and I will add film caps to the incoming signal cap C5010, add film caps decoupling C5130 and 5120.
The ones I need some help/advice with are C5070 and 5080, they togeather with C5010 are audio grade caps, 5070 and 5080 are Nichi Gold caps so I am assuming they would be important.
Do you agree adding film caps to C5070 and 5080 will improve things or would I mess up the driver amp stage?
I've done the D/A section and low level analog path as well as the power end of the amp section and have started to look at the driver amp section to see if there is potential. The mods are generally just get rid of as many caps is the signalpath as you can, if you need them add a film cap and beef up powersupply and decouple with film caps. Finally make sure the grounding is solid. In the picture you see the driver amp section and I will add film caps to the incoming signal cap C5010, add film caps decoupling C5130 and 5120.
The ones I need some help/advice with are C5070 and 5080, they togeather with C5010 are audio grade caps, 5070 and 5080 are Nichi Gold caps so I am assuming they would be important.
Do you agree adding film caps to C5070 and 5080 will improve things or would I mess up the driver amp stage?
Attachments
if C5070 & C5080 don't have a plus on one end, I wouldn't monkey with them. If they are electrolytic with a plus, I would audio test the circuit before doing repairs or mods.
Start out, listen to a real grand piano, Yamaha or Steinway either one. Their sound is close. Listen to all octaves, especially the top octave.
Then listen to your amp on good $50 headphones. I own some Koss ones. Does the amp sound the same? If so, no problem so far. MP3 is useless IMHO for this test, good piano CDs and LP's come from Colombia (not special products), RCA red seal or dynagroove (not camden), Telefunken, Telearc, Phillips a few others. Mst be a post 1965 recording to have condensor mikes, and RCA used ribbon mikes up to about 1969 (bad).
Then listen to the amp on your speakers. does the amp and speakers sound the same? If not your speakers probably at fault.
Another cheap amp check is play a compressed pop track, loud as you can stand it before it sounds awful (clipping) . Measure the AC voltage on the speakers at that level. Use an analog VOM to measure average voltage, or using a scope (multiply peak to peak voltage by .7). P = (1/Z)*V^2 where Z is speaker impedance and P is power. Is your amp putting out the advertised power? Tired amps put out less than the advertised power.
If your amp can do pianos accurately, and puts out the advertised power, spend your money on your speakers. I did.
If the power is low, then determine the age of the amp. Over 15 years, you can measure the electrolytic caps that tend to go first, with a $100 Peak capacitance/ESR meter. Or, if I determine the amp is at fault with the piano test or the power test, i start replacing e-caps with premium long life versions (which have been excellent so far, without paying for "audio" snake oil capacitors.) Long life is >3000 hours for 1500 uf or bigger, >8000 hours for smaller ones. Top grade e-cap suppliers have been panasonic, nichicon, rubicon, vishay. Read the service life spec, they all sell trash if the purchaser really wants it.
I don't monkey with film caps in amps unless they were pre-1965 (paper) or rail voltage blasted through the driver boards due to a shorted output transistor, blowing up parts.
I will say the Onkyo has +-65 v rails, it should have some spectacular power peaks on media with a lot of dynamic range. The single output transistor pair, however, should limit constant power out to 50 W./ch or less. They would have to TO3 case or those sankyo 3 cm wide plastic transistors to produce even that. TO66 or TO220, more like 25 W/ch. In the receiver market, usually a lot less than 50 or 25 W due to weight limitation and competitive pricing limiting heat sink and transformer size.
Just testing is a good intro to diyaudio without actually breaking things with the soldering iron. Onkyo has a good reputation, pick up a junky Sony or panasonic or akai or kenwood or sherwood to mess with before touching this. IMHO. Or pick up a road weary PA amp; without the reciever they are easier to open up and close back. I started transistor amps with a ****y dynaco ST120, the "worst amp ever made" according to some. Sounds really good now, that I've done with it.
Start out, listen to a real grand piano, Yamaha or Steinway either one. Their sound is close. Listen to all octaves, especially the top octave.
Then listen to your amp on good $50 headphones. I own some Koss ones. Does the amp sound the same? If so, no problem so far. MP3 is useless IMHO for this test, good piano CDs and LP's come from Colombia (not special products), RCA red seal or dynagroove (not camden), Telefunken, Telearc, Phillips a few others. Mst be a post 1965 recording to have condensor mikes, and RCA used ribbon mikes up to about 1969 (bad).
Then listen to the amp on your speakers. does the amp and speakers sound the same? If not your speakers probably at fault.
Another cheap amp check is play a compressed pop track, loud as you can stand it before it sounds awful (clipping) . Measure the AC voltage on the speakers at that level. Use an analog VOM to measure average voltage, or using a scope (multiply peak to peak voltage by .7). P = (1/Z)*V^2 where Z is speaker impedance and P is power. Is your amp putting out the advertised power? Tired amps put out less than the advertised power.
If your amp can do pianos accurately, and puts out the advertised power, spend your money on your speakers. I did.
If the power is low, then determine the age of the amp. Over 15 years, you can measure the electrolytic caps that tend to go first, with a $100 Peak capacitance/ESR meter. Or, if I determine the amp is at fault with the piano test or the power test, i start replacing e-caps with premium long life versions (which have been excellent so far, without paying for "audio" snake oil capacitors.) Long life is >3000 hours for 1500 uf or bigger, >8000 hours for smaller ones. Top grade e-cap suppliers have been panasonic, nichicon, rubicon, vishay. Read the service life spec, they all sell trash if the purchaser really wants it.
I don't monkey with film caps in amps unless they were pre-1965 (paper) or rail voltage blasted through the driver boards due to a shorted output transistor, blowing up parts.
I will say the Onkyo has +-65 v rails, it should have some spectacular power peaks on media with a lot of dynamic range. The single output transistor pair, however, should limit constant power out to 50 W./ch or less. They would have to TO3 case or those sankyo 3 cm wide plastic transistors to produce even that. TO66 or TO220, more like 25 W/ch. In the receiver market, usually a lot less than 50 or 25 W due to weight limitation and competitive pricing limiting heat sink and transformer size.
Just testing is a good intro to diyaudio without actually breaking things with the soldering iron. Onkyo has a good reputation, pick up a junky Sony or panasonic or akai or kenwood or sherwood to mess with before touching this. IMHO. Or pick up a road weary PA amp; without the reciever they are easier to open up and close back. I started transistor amps with a ****y dynaco ST120, the "worst amp ever made" according to some. Sounds really good now, that I've done with it.
Last edited:
It depends on what you mean by improvements. Just applying very generalised advice like bypassing electolytics with film caps or omitting them, may or may not apply to the line level signals in your circuits. You could just as easily claim that adding superfluous components increases noise susceptibility or that omitting signal caps introduces DC currents and offset problems.
I tend to think that the power amplifier(s) are best left in their present format. Look elsewhere to make "improvements" but if you like to feel you are taking action, look at the DC potential across C5010, for example. If it is below 1-2V, the cap will probably not be polarized sufficiently and may no longer be working properly. That can be remedied by using good quality NP types like Nichicon ES. Try that first in a couple of channels that can be used as a stereo or 2 x mono pair, if physical dimensions permit.
I tend to think that the power amplifier(s) are best left in their present format. Look elsewhere to make "improvements" but if you like to feel you are taking action, look at the DC potential across C5010, for example. If it is below 1-2V, the cap will probably not be polarized sufficiently and may no longer be working properly. That can be remedied by using good quality NP types like Nichicon ES. Try that first in a couple of channels that can be used as a stereo or 2 x mono pair, if physical dimensions permit.
Member
Joined 2009
Paid Member
I improved my modern Pioneer HT amplifier by modifying the amplifier. It wasn't a great design. Your Onkyo amp looks much better already in terms of the amplifier topology. I doubt there's anything worth doing with it. Even electrolytic caps in the signal path will likely be fine, film caps will do nothing, you could try Nichicon audio bipolar caps if you must.
I found the main issue was source quality - most of the movies I play simply don't have great sound tracks. Some of them use DTS(?) instead of MP3 and are far better. My favourite sound track is from The Master and the Commander.
I found the main issue was source quality - most of the movies I play simply don't have great sound tracks. Some of them use DTS(?) instead of MP3 and are far better. My favourite sound track is from The Master and the Commander.
If Onkyo needed small film HF blocking across the decoupling capacitors, they would have fitted them.
Don't waste your money on useless Snake Oil filled capacitors. Back in the 60s we used to replace those with more reliable, (new fangled), polythene or PRP caps. Maybe the reason the market is flooded with oil filled capacitors is that they didn't sell.
Good for motor start and fluorescent lamp ballast uses though.
Don't waste your money on useless Snake Oil filled capacitors. Back in the 60s we used to replace those with more reliable, (new fangled), polythene or PRP caps. Maybe the reason the market is flooded with oil filled capacitors is that they didn't sell.
Good for motor start and fluorescent lamp ballast uses though.
Things have not changed on here I notice, I used to do this a lot many years ago and everytime one asked for advice one usually got a lot of answers to questions not asked.
Either way I should not have written improved, I agree that the changes I make don't change any measurable attribute of the amp. Perhaps modify is better and what I am trying to get to is a sonic signature that I enjoy if you will. I like them transparent, powerful and clean.
The amp has always been amazingly powerful (there are 4*80W power trans per channel for all channels and a 1kw trafo) and generally very good but a bit soft. Lately its also gotten a bit grainy for the first 1-2 h of use and as it gets really hot and all the caps are std 1000h 85 rated I have started to think the caps have degraded and I have nothing to loose by modifying it a bit, it would have been scrapped otherwise.
In general and in my experiance modding maybe 20 amps/CD/DVD/DA/Preamps etc over the years the modifications I am considering have always changed things to the better so there is no doubt in my mind that it helps and I would really appriciate some insights on what the C5070/80 are doing there and what to consider if I make changes to them.
They are polarized btw and thanks to those of you who are adressing my question on these cap's.
Either way I should not have written improved, I agree that the changes I make don't change any measurable attribute of the amp. Perhaps modify is better and what I am trying to get to is a sonic signature that I enjoy if you will. I like them transparent, powerful and clean.
The amp has always been amazingly powerful (there are 4*80W power trans per channel for all channels and a 1kw trafo) and generally very good but a bit soft. Lately its also gotten a bit grainy for the first 1-2 h of use and as it gets really hot and all the caps are std 1000h 85 rated I have started to think the caps have degraded and I have nothing to loose by modifying it a bit, it would have been scrapped otherwise.
In general and in my experiance modding maybe 20 amps/CD/DVD/DA/Preamps etc over the years the modifications I am considering have always changed things to the better so there is no doubt in my mind that it helps and I would really appriciate some insights on what the C5070/80 are doing there and what to consider if I make changes to them.
They are polarized btw and thanks to those of you who are adressing my question on these cap's.
C5070 and C5080 do a little frequency modification to the VAS stage. A little more bass I think, C5010 is the one that will really mess up your sound if old. The only tantalums I've bought I used for input caps, but the ones I bought came with popcorn noise. So I've been using 4.7 uf 50v aerovox gold CGO ceramic caps for that position, Very nice sound, the excess voltage (50v rating on 2 v signal) must linearize them some.
Really gritty sound is why I change aged electrolytics. You especially notice it as the room cools down in the winter.
On the rail caps, wattage out I find a better discriminator than ESR for bad ones. A 100 W organ amp downtown from 1980 quit working one Sunday in the middle of the service. The next Monday I found it working cold but putting out 2 W. I've since measered these 14000 uf caps with a Peak ESR meter and the two worst ones are 14500 uf at .82 ohms and 16100 uf at .17 ohms. Those numbers do not sound awful, but 2 watts out of 100 sure does.
Really gritty sound is why I change aged electrolytics. You especially notice it as the room cools down in the winter.
On the rail caps, wattage out I find a better discriminator than ESR for bad ones. A 100 W organ amp downtown from 1980 quit working one Sunday in the middle of the service. The next Monday I found it working cold but putting out 2 W. I've since measered these 14000 uf caps with a Peak ESR meter and the two worst ones are 14500 uf at .82 ohms and 16100 uf at .17 ohms. Those numbers do not sound awful, but 2 watts out of 100 sure does.
Whether you can go 1k depends on the Vceo, wattage, SOA, and heat sink capability of the transistors in your amp. Same part number, but your amp has higher rail voltages. You'll have to download the datasheet and calculate. You haven't seen the heat sink in the other amp to know what they use. If you read a lot of repair threads here, people find drivers & occasionally VAS (voltage amp stage) transistors toasted. I put cheapo heat sinks on all my VAS & drivers in custom builds instead of doing a lot of calculating. Another caveat, your amp may have selected high gain drivers of the same part number so it doesn't need the high drive current of a 1k collector resistor.
The fear of capacitors is something I can't quite understand. Humans like a little phase shift in their music, phas shift is what effect a room has on a speaker output. Bands sometimes put phase shift in with digital effect boxes these days to pep up their sound. Yes ecaps and tantalums go bad. I've spent $3.68 in some cases replacing dried up 10 uf e-caps in a interstage coupler with film caps so I don't have to replace them again. But I didn't calculate the operating points to replace them with resistors and direct couple the stages.
The 10 uf you could probably replace with a 10 uf roederstein film cap: watch the 100 v rating. . The 22 you'd need two 10 uf parallel at $3.63 each.
It is possible to do a frequency sweep of an amp using PC's & sound cards. I'd cut the frequency mod e-caps one leg and do that before & after, or a good full keyboard piano sound check of the response, before leaving the caps out. & make sure the power supply rail caps ares good before checking these. Dried up power supply rail caps affect first octave bass more than treble. The rail caps can run out of energy before a 32 hz note has gotten half way through. My organs had a lot of that bass loss when I bought them - the reason they were $$ instead of $$$$.
The fear of capacitors is something I can't quite understand. Humans like a little phase shift in their music, phas shift is what effect a room has on a speaker output. Bands sometimes put phase shift in with digital effect boxes these days to pep up their sound. Yes ecaps and tantalums go bad. I've spent $3.68 in some cases replacing dried up 10 uf e-caps in a interstage coupler with film caps so I don't have to replace them again. But I didn't calculate the operating points to replace them with resistors and direct couple the stages.
The 10 uf you could probably replace with a 10 uf roederstein film cap: watch the 100 v rating. . The 22 you'd need two 10 uf parallel at $3.63 each.
It is possible to do a frequency sweep of an amp using PC's & sound cards. I'd cut the frequency mod e-caps one leg and do that before & after, or a good full keyboard piano sound check of the response, before leaving the caps out. & make sure the power supply rail caps ares good before checking these. Dried up power supply rail caps affect first octave bass more than treble. The rail caps can run out of energy before a 32 hz note has gotten half way through. My organs had a lot of that bass loss when I bought them - the reason they were $$ instead of $$$$.
Last edited:
Sharing my findings if someone else wants to make changes to their amp stage. I decided to add film caps to C5070/80 and 5130/20 (C5010 was already done). I did it in steps and the 5070/80 opened up the soundstage without any other apparent negative side effect. However, adding film caps to 5130/20 messed up the high frequencies making them less defined and slightly distorted.
Just to be clear I had done a lot of mods before getting to this. I was chasing the last bit of graininess which in my experience usually comes from older elyt cap in the signal path.
I now have a full film cap signal path and with diverse powersupply fixes from hdmi in and onwards and now have a very clean, spacial, dynamic and powerful sound stage. It used to be clean, a bit muffled albeit always powerful.
Everything else beeig equal I can now crank it up another 6dB without thinking its to loud.
Just to be clear I had done a lot of mods before getting to this. I was chasing the last bit of graininess which in my experience usually comes from older elyt cap in the signal path.
I now have a full film cap signal path and with diverse powersupply fixes from hdmi in and onwards and now have a very clean, spacial, dynamic and powerful sound stage. It used to be clean, a bit muffled albeit always powerful.
Everything else beeig equal I can now crank it up another 6dB without thinking its to loud.
IF the amp is overage with leaky seals on the electrolytic caps, then paralleling them with a film cap just postpones the inevitable deterioration of sound. It might continue. This is a case where 1 I try a new long life e-cap in the same position or 2 shotgun replace it with a film cap removing the old e-cap, or 3 measure the ESR and compare to a new cap. (new for 2016, a Peak ESR meter)
These old organ amps and preamps, I don't need a blind test setup to tell the difference between old & new caps. I put a couple new e-caps in replacing the old, button it up, most of the time it sounds better on one function. I have 6 preamps for different functions in one organ model, so a replaced e-cap only affects that function. The 42 year old hifi amp I repaired, the second new input cap made a big differnce in sound, the rail and out caps didn't change anything audible. Changed them anyway. First new input caps (tantalum) had popcorn noise, last time I buy parts from that TV repair supply store.
Rather than butcher the circuit board with solder sucker etc, possibly pulling lands, I usually just cut one lead on the old cap, make hooks, and hook the new cap over the leads of the old one. One lead cut, deterioration doesn't affect the sound.
Glad you like your amp now.
These old organ amps and preamps, I don't need a blind test setup to tell the difference between old & new caps. I put a couple new e-caps in replacing the old, button it up, most of the time it sounds better on one function. I have 6 preamps for different functions in one organ model, so a replaced e-cap only affects that function. The 42 year old hifi amp I repaired, the second new input cap made a big differnce in sound, the rail and out caps didn't change anything audible. Changed them anyway. First new input caps (tantalum) had popcorn noise, last time I buy parts from that TV repair supply store.
Rather than butcher the circuit board with solder sucker etc, possibly pulling lands, I usually just cut one lead on the old cap, make hooks, and hook the new cap over the leads of the old one. One lead cut, deterioration doesn't affect the sound.
Glad you like your amp now.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.