"I don't have the feeling that this amps sings. It just WORKS (and quite reliably BTW)."
I agree Charles. 30 years later and this amp is still functional... and I'm sure it was abused and mishandled (judging by the dents and scratches). None the less, from the moment I bought it and powered it up she still had some get up and go. SQ is not good (and I am by no means and audiophile with a trained ear) but the power is there.
Assuming that Darkenfriz knows his stuff (which he sure seems to) and this is a "Type AB + C..." amp then there are fundamental issues that I will never correct through bias settings and audio grade components. However; as I am using this amp for bass only I can probably afford to live with a less than perfect amplifier... so long as it has some grunt (ie. power) behind it.
As a starting point:
I know that electrolytics don't last forever, so I will replace the PSU caps and signal line electro's. While I'm at it I might as well add the snubbing caps.
The volume control pots are badly corroded. I had to disassmble the pots and scrape corrosion off of the contacts so that the amp would function properly. The pots will therefore be replaced.
Perhaps a high quality Op-amp will improve things a bit. Maybe it won't. An opamp only costs a few bucks... what's the harm in giving it a shot?
After all of these upgrades are complete and if I am still not satisfied with this amp then I am left with big heatsinks a rigid power supply and a solid case. Sounds like excellent starting point for a DIY amp project. I contacted Greg Ball and he tells me that the GB300D will operate on 81V rails. A simple matter to convert this commercial beast into a wicked HiFi High power unit. Or perhaps I could purchase a Class D module... Insane power for a sealed subwoofer project with Linkwitz transform? A class D amp on these rails would have near 600Wrms... (Is this correct?).
All in all, none of the upgrades will go to waste regardles of the path i take with this amp.
I agree Charles. 30 years later and this amp is still functional... and I'm sure it was abused and mishandled (judging by the dents and scratches). None the less, from the moment I bought it and powered it up she still had some get up and go. SQ is not good (and I am by no means and audiophile with a trained ear) but the power is there.
Assuming that Darkenfriz knows his stuff (which he sure seems to) and this is a "Type AB + C..." amp then there are fundamental issues that I will never correct through bias settings and audio grade components. However; as I am using this amp for bass only I can probably afford to live with a less than perfect amplifier... so long as it has some grunt (ie. power) behind it.
As a starting point:
I know that electrolytics don't last forever, so I will replace the PSU caps and signal line electro's. While I'm at it I might as well add the snubbing caps.
The volume control pots are badly corroded. I had to disassmble the pots and scrape corrosion off of the contacts so that the amp would function properly. The pots will therefore be replaced.
Perhaps a high quality Op-amp will improve things a bit. Maybe it won't. An opamp only costs a few bucks... what's the harm in giving it a shot?
After all of these upgrades are complete and if I am still not satisfied with this amp then I am left with big heatsinks a rigid power supply and a solid case. Sounds like excellent starting point for a DIY amp project. I contacted Greg Ball and he tells me that the GB300D will operate on 81V rails. A simple matter to convert this commercial beast into a wicked HiFi High power unit. Or perhaps I could purchase a Class D module... Insane power for a sealed subwoofer project with Linkwitz transform? A class D amp on these rails would have near 600Wrms... (Is this correct?).
All in all, none of the upgrades will go to waste regardles of the path i take with this amp.
I have an 8.5C and a TNT bass amp that I am happy with. Peavey stuff is OK, but the pros moan about it. As a learning project its worth playing with, but if it is to USE consider this:-
Peavey never were amaturs. My guess it that they tweeked the hell out of their hardware at the design stage. - Watts sell amps. They really do sell bang for buck gear.
If you upped performance by 10% so what? 80W more is 0.4db. You shouldn't be running it hard enough to notice.
For the money that you would spend making it run you could get somthing good on Ebay.
That said, It worth breaking it to learn on.
Peavey never were amaturs. My guess it that they tweeked the hell out of their hardware at the design stage. - Watts sell amps. They really do sell bang for buck gear.
If you upped performance by 10% so what? 80W more is 0.4db. You shouldn't be running it hard enough to notice.
For the money that you would spend making it run you could get somthing good on Ebay.
That said, It worth breaking it to learn on.
" ... I don't think that every developer was aware of such details back then. ..."
Peavey has always been considered the bottom rung on the quality ladder, but high up the reliability ladder ... though not necessarily justified in either requirement. The Peavey engineering of the seventies and eighties was not the best ...
... and the MKT / snubbing cap trick was not very well understood then. Now that it is SOP for Analog Devices and Texas Instrument engineers to include this trick in their tech reports on analog amps and op-amps, it has begun to trickle down to the service level.
Notice that this op-amp is recommended to be installed with snubbing caps, right there on page 8 of the data sheet: http://www.analog.com/UploadedFiles/Data_Sheets/AD825.pdf ... note that there are two sets of caps on the power pins, electrolytic and plastic MKT types (10 uF and 0.01 uF respectively). This is one of the main reasons that modern op-amps and amplifiers are able to get distortion and phasing under control = those plastic MKT snubbing caps make these puppies sing = producing the graphis on page 7, etc., and on just about every other page that has a diagram = distortion numbers better than -85 db and common mode rejection per channel greater than 100 db in the audio range = dramatically better than the numbers for high quality op-amps of the seventies and eighties.
... AND adding plastic snubbing caps to high power amplifiers, even retro quality produces similar results ... makes 'em sing much sweeter ...
Peavey has always been considered the bottom rung on the quality ladder, but high up the reliability ladder ... though not necessarily justified in either requirement. The Peavey engineering of the seventies and eighties was not the best ...
... and the MKT / snubbing cap trick was not very well understood then. Now that it is SOP for Analog Devices and Texas Instrument engineers to include this trick in their tech reports on analog amps and op-amps, it has begun to trickle down to the service level.
Notice that this op-amp is recommended to be installed with snubbing caps, right there on page 8 of the data sheet: http://www.analog.com/UploadedFiles/Data_Sheets/AD825.pdf ... note that there are two sets of caps on the power pins, electrolytic and plastic MKT types (10 uF and 0.01 uF respectively). This is one of the main reasons that modern op-amps and amplifiers are able to get distortion and phasing under control = those plastic MKT snubbing caps make these puppies sing = producing the graphis on page 7, etc., and on just about every other page that has a diagram = distortion numbers better than -85 db and common mode rejection per channel greater than 100 db in the audio range = dramatically better than the numbers for high quality op-amps of the seventies and eighties.
... AND adding plastic snubbing caps to high power amplifiers, even retro quality produces similar results ... makes 'em sing much sweeter ...
" ... I'm an analog and measurement guy. If somebody says some audio circuits don't sound the same, I bet I can measure the difference. I thought of the old claim (by people with "good ears") that electrolytic capacitors in C-R coupling networks don't sound the same as high-quality film capacitors in an ordinary audio circuit. ..." From: http://www.elecdesign.com/Articles/ArticleID/13166/13166.html
More like that: Tweaking caps and wires on the bench to fudge and fix the results:
http://www.elecdesign.com/Articles/ArticleID/14109/14109.html
http://www.elecdesign.com/Articles/ArticleID/15116/15116.html
More like that: Tweaking caps and wires on the bench to fudge and fix the results:
http://www.elecdesign.com/Articles/ArticleID/14109/14109.html
http://www.elecdesign.com/Articles/ArticleID/15116/15116.html
FastEddy said:
The MKT capacitors referenced above may be designed as snubber caps. Generally snubbers are used in applications where switching power transistors need DVDT protection. I did not see in the app note where the MKT caps are specified. All they call out in the text are the standard .10 mfd bypass capacitors. The schematics do show a 10uf and a .01uf bypass cap. I did not see anything that specified a particular cap.
" ... I did not see in the app note where the MKT caps are specified. ... The schematics do show a 10uf and a .01uf bypass cap. I did not see anything that specified a particular cap. ..."
Neither do I. That's because they don't like to give away their performance secrets to the competition. If you check out the oriental analog chip makers, you will notice that they are not hip to this trick, yet.
The schemos do indicate that the 10 uF (@ the op-amp power pins) is an electrolytic ... the other remains unknown, except to those in the know. But if you ever go to an IEEE conference that has to do with power supplies or analog gadgets or amplification and just ask, you get the message ... but you gotta know what to ask for ...
Neither do I. That's because they don't like to give away their performance secrets to the competition. If you check out the oriental analog chip makers, you will notice that they are not hip to this trick, yet.
The schemos do indicate that the 10 uF (@ the op-amp power pins) is an electrolytic ... the other remains unknown, except to those in the know. But if you ever go to an IEEE conference that has to do with power supplies or analog gadgets or amplification and just ask, you get the message ... but you gotta know what to ask for ...
I stated earlier that adjusting bias is probably not worth the effort for this amp. Am I wrong? Has anyone reviewed the schematic, is it possible to increase the bias? If anyone is confident that they understand exactly how the bias is set for this amp then I would be happy to give it a shot.
I have not received much feedback regarding upgrading the opamp to a Burr-Brown OPA134. This is simple enough and I can add snub caps while I'm at it. But is there any concern that I could wind up getting into oscillation problems with this hi-speed device? Is there anything else I'm missing? Am I oversimplifying things? Can the OPA134 simply be soldered in place of the existing unit?
I have not received much feedback regarding upgrading the opamp to a Burr-Brown OPA134. This is simple enough and I can add snub caps while I'm at it. But is there any concern that I could wind up getting into oscillation problems with this hi-speed device? Is there anything else I'm missing? Am I oversimplifying things? Can the OPA134 simply be soldered in place of the existing unit?
Charles,
If there is no Bias current then adding bias would improve the SQ... No? Would it be possible to apply a bias to the output devices? Is there no way to convert this amp to class AB? I'm starting to go in circles here, I've already asked this question but didn't receive a definitive response.
If there is no Bias current then adding bias would improve the SQ... No? Would it be possible to apply a bias to the output devices? Is there no way to convert this amp to class AB? I'm starting to go in circles here, I've already asked this question but didn't receive a definitive response.
I think, that easiest and most safe way of adding some bias to power transistors is to swap R19 (100R) with a NTC thermistor.
Mount thermistor on main heatsink or glue it to output transistor, buy yourself several and check how bias behaves, 330ohms at cold may be a good start value.
regards
Adam
Mount thermistor on main heatsink or glue it to output transistor, buy yourself several and check how bias behaves, 330ohms at cold may be a good start value.
regards
Adam
Thanks for the input Adam. I know you said before that besides PSU caps there is not much worth tweaking in this amp. Do you feel that taking these steps towards adding bias is a waste of time? Also, are you quite confident in the technique that you suggested, I would prefer not to roast my amp if there is much more than a shadow of doubt in your mind as to how it functions. (This is not meant to sound rude, I understand that I asked your opinion and you offered it)
Of course you risk frying this and that and results are doubttful, maybe it will sound some better, maybe not, maybe just a little.
I meant if you insist on adding some hi-fi bias I suggest doing it this way.
You must decide if you want to tweak, I cannot fully predict the results.
BTW, if there was no risk of frying, than tweaking would be much less fun.
I meant if you insist on adding some hi-fi bias I suggest doing it this way.
You must decide if you want to tweak, I cannot fully predict the results.
BTW, if there was no risk of frying, than tweaking would be much less fun.
the bias is set by the dual diodes in the driver section. you could add a little bias by adding another diode stack in the string, but it's not recommended, since an over-biased cs-800 is a suitcase nuke waiting to happen
.........
it's difficult enough rebuilding one that's had somebody trip over the speaker cord onstage, and the amp spends the rest of the evening driving a short........
the bias stack biases the drivers, which bias the output devices to just at conduction. as you will notice from the schematic, they give you the voltage across the driver emitter resistor. that's because at idle, the drivers are drawing some current, and the output devices are just barely below conduction. you could add some bias with another diode or two, but if you let the magic smoke out, i will disavow any knowledge of your actions........
.........it's difficult enough rebuilding one that's had somebody trip over the speaker cord onstage, and the amp spends the rest of the evening driving a short........
the bias stack biases the drivers, which bias the output devices to just at conduction. as you will notice from the schematic, they give you the voltage across the driver emitter resistor. that's because at idle, the drivers are drawing some current, and the output devices are just barely below conduction. you could add some bias with another diode or two, but if you let the magic smoke out, i will disavow any knowledge of your actions........
Well, fyi297 I don't see your point.
As long as somebody has dredged this up, the best modification available to a CS800b is to remove the 1/4 phone jack outputs (per the problem unclejed613 points out) and replace them with nutrix connectors. Or euro terminal strip and bare wires on the speakers. Or anything else but 1/4 phone jack with its instant short when pulled out 1/8".
I've got the oversized too much hormones son of CS800b, the PV-1.3k, and having replaced over 100 parts after a meltdown, (I bought it junk) wonder what I can do to get 20 ma across the emitter resistors, cold. A whole diode in the bias stack is way too much. I've been thinking of changing the resistor across the non-heatsink diode in the stack, but a thermistor on the heatsink sounds way cooler. Unfortunately, the PWB wiring is such that it would require cutting the trace and drilling some holes, and a broken circuit diode stack is a trigger to a core meltdown for sure. I was kind of wondering if one transistor is conducting 20 ma (since I can only measure one pair of the 5 emitter resistors per channel, and that only because I installed two terminal strips to make them accessible). The transistor pair I have the terminal strip attached to, on each channel, has a cold bias current, of zero. Since my intended use is classical music in the living room, not pounding techo-rock at the disco, a little bias current might be really nice. Yeah, I should have bought a McIntosh, but this was $55, and the mac was $1000 junk. Even a lowly dynaco ST-400 is $400 unrestored. This is diyaudio, after all; this bull of an amp could use a little refinement to go with the almost bulletproof design.
As long as somebody has dredged this up, the best modification available to a CS800b is to remove the 1/4 phone jack outputs (per the problem unclejed613 points out) and replace them with nutrix connectors. Or euro terminal strip and bare wires on the speakers. Or anything else but 1/4 phone jack with its instant short when pulled out 1/8".
I've got the oversized too much hormones son of CS800b, the PV-1.3k, and having replaced over 100 parts after a meltdown, (I bought it junk) wonder what I can do to get 20 ma across the emitter resistors, cold. A whole diode in the bias stack is way too much. I've been thinking of changing the resistor across the non-heatsink diode in the stack, but a thermistor on the heatsink sounds way cooler. Unfortunately, the PWB wiring is such that it would require cutting the trace and drilling some holes, and a broken circuit diode stack is a trigger to a core meltdown for sure. I was kind of wondering if one transistor is conducting 20 ma (since I can only measure one pair of the 5 emitter resistors per channel, and that only because I installed two terminal strips to make them accessible). The transistor pair I have the terminal strip attached to, on each channel, has a cold bias current, of zero. Since my intended use is classical music in the living room, not pounding techo-rock at the disco, a little bias current might be really nice. Yeah, I should have bought a McIntosh, but this was $55, and the mac was $1000 junk. Even a lowly dynaco ST-400 is $400 unrestored. This is diyaudio, after all; this bull of an amp could use a little refinement to go with the almost bulletproof design.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.