Hi all. Medium-duration lurker, new poster.
I've got a Peavey CS-800x in good, functional condition. A little rough cosmetically, but is well-behaved. I've owned it for about 15 years -- used it in the stereotypical application of driving a couple of DJ subwoofers for a small mobile rig when I was in high school.
I've held onto it because I always regarded it as a nice piece of gear. As I turn my attention toward two-channel home audio, it's occurred to me that it might have a place in one of my systems around the house. My first inclination is to make it part of my basement system, driving a couple of HPM-100s for the workshop/gym.
I have read everything I can find on the CS-800 variants for home stereo use, and I'm left with a couple of questions:
I've got a Peavey CS-800x in good, functional condition. A little rough cosmetically, but is well-behaved. I've owned it for about 15 years -- used it in the stereotypical application of driving a couple of DJ subwoofers for a small mobile rig when I was in high school.
I've held onto it because I always regarded it as a nice piece of gear. As I turn my attention toward two-channel home audio, it's occurred to me that it might have a place in one of my systems around the house. My first inclination is to make it part of my basement system, driving a couple of HPM-100s for the workshop/gym.
I have read everything I can find on the CS-800 variants for home stereo use, and I'm left with a couple of questions:
- Opinions seem divided on whether the CS-800 variants have the sound quality chops that make them suitable for home use. Is this only applicable to cases where it's being driven hard, or...? On paper, the specs appear very good. My hope is that it would make a great amp to drive a couple of 8ohm/100W or 4ohm/200W speakers with headroom to spare. Am I off-base?
- I assume that it would benefit from some maintenance. I've already done a de-oxit pass, and I've checked that I have no excessive DC offset (7mV on A / 0.8mV on B). Would capacitor replacement be an easy-enough process for a novice on this amp? I'm an engineer, but mechanical - not electrical. I'm good with a soldering iron, but have not done any work on amplifiers before.
- I've got a bit of hiss out of both channels. Only audible when close to the speakers, and when no sound is playing. Is this just a product of a higher noise floor in the 80's op-amps, or might it indicate some issues that I should correct?
It seems to be in good shape, and adequate for a basement system.
Use it for a while, and don't do any more work on it unless necessary,
except if you want to use it as a learning experience. The hiss is common,
especially in high gain circuits like that.
Use it for a while, and don't do any more work on it unless necessary,
except if you want to use it as a learning experience. The hiss is common,
especially in high gain circuits like that.
Last edited:
I find my CS800s which has about the same driver circuitry as cs800x but different power supply, very refined at 1 W/ch in my living room. I had 102 db 1w1m speakers (sp2-xt) until September 2020, and they would have revealed any hiss. My cs800s sounded as good as my modified djoffe bias control dynakit ST120 on a good day, and the cs800s is still sounding good whereas the modified ST120 popped a solder joint and hums now. The cs800s never popped loudly during lightning storms or anything either. I never had loose cables that would cause big pops from a 400 w/ch amp though. Don't cheap out on dodgy cables in a living room with a high watt amp, without earplugs.
You can do a little op amp swapping to improve on the hiss maybe. My cs800s has njm4560 for U100 which is very refined. I can't exactly read U100 on the cs800x but it may be RA3678 which according to electronicsclub.cjb.net is either 4560 or 5532. Any 10k or up carbon film resistors like R117 could be upgraded to metal film, too. Metal film hisses a little less in higher resistances.
I've found the 2.2 uf NP input cap a bit of a bother in various 40 year old peavey amps. Not shown on the driver board schematics: it is located on the input board on the back next to the volume control. I use 2.2 uf ceramic caps there: COG if I can get it but 50 v rated X7R in any case. 2 v or less input signal on 50 v rated ceramic is not all that non-linear, IMHO. I haven't detected any microphonics but I don't do high watt stage shows with the amp sitting under the bass bins.
The 10000 uf mains caps in my PV-1.3k were a little high esr, but that would be a problem only at 650 w/ch. In my living room, no problem. I changed them to triple parallel 3300 uf caps, but I don't play big enough audiences to need that many watts.
The PV-4c the full volume watts were seriously cramped by dried up mains caps.
The cs800s blew a 47 uf cap in the switcher supply, and the fuse in front of it, so I changed all the caps in that sub-assembly on general principal. Old caps are bought to roughly the same life rating, so when one fails, they all are scheduled to go and the amp would be broken all the time if I let them fail one at a time. Peavey rail caps don't usually fail & leak but if used until hot they go high ESR and crimp the full volume wattage in their old age. Full volume wattage can be tested on 4 ohm resistors of appropriate wattage, with an analog AC voltmeter or a scope. P=(V^2)/Z where Z is load impedance. Most 90's Peavey products are power rated at 4 ohms load.
Happy hot-rodding.
You can do a little op amp swapping to improve on the hiss maybe. My cs800s has njm4560 for U100 which is very refined. I can't exactly read U100 on the cs800x but it may be RA3678 which according to electronicsclub.cjb.net is either 4560 or 5532. Any 10k or up carbon film resistors like R117 could be upgraded to metal film, too. Metal film hisses a little less in higher resistances.
I've found the 2.2 uf NP input cap a bit of a bother in various 40 year old peavey amps. Not shown on the driver board schematics: it is located on the input board on the back next to the volume control. I use 2.2 uf ceramic caps there: COG if I can get it but 50 v rated X7R in any case. 2 v or less input signal on 50 v rated ceramic is not all that non-linear, IMHO. I haven't detected any microphonics but I don't do high watt stage shows with the amp sitting under the bass bins.
The 10000 uf mains caps in my PV-1.3k were a little high esr, but that would be a problem only at 650 w/ch. In my living room, no problem. I changed them to triple parallel 3300 uf caps, but I don't play big enough audiences to need that many watts.
The PV-4c the full volume watts were seriously cramped by dried up mains caps.
The cs800s blew a 47 uf cap in the switcher supply, and the fuse in front of it, so I changed all the caps in that sub-assembly on general principal. Old caps are bought to roughly the same life rating, so when one fails, they all are scheduled to go and the amp would be broken all the time if I let them fail one at a time. Peavey rail caps don't usually fail & leak but if used until hot they go high ESR and crimp the full volume wattage in their old age. Full volume wattage can be tested on 4 ohm resistors of appropriate wattage, with an analog AC voltmeter or a scope. P=(V^2)/Z where Z is load impedance. Most 90's Peavey products are power rated at 4 ohms load.
Happy hot-rodding.
Last edited:
Use a film cap for the 2.2 uf inch input, polypropelene if you can fit it, but at least polyester
it appears to be noisy at 100 db s/n from 840W.
my amp is 115db from 100W. ( 95db at 1W) I can barely hear any hiss with my ear pressed to the tweeter.
my amp is 115db from 100W. ( 95db at 1W) I can barely hear any hiss with my ear pressed to the tweeter.
PA amps often don't care about noise levels. Its very common for PA systems to have loud hiss (by domestic standards) due to a combination of high driver efficiencies and low noise being ignored during circuit design. And of course the cooling fan will make 40dB or more noise anyway.
Using lower sensitivity speakers can help mitigate the issue.
Using lower sensitivity speakers can help mitigate the issue.
There is a radical difference in hiss between the CS800 rev A to C which used a uA741 op amp, and the CS800x & s. I find my cs800s hiss nearly inaudible. I use 102 db 1w1m speakers.
The fan doesn't run at low wattage in the s. It is temperature controlled. I
Most of those old PA amps (including the CS800) that used 741 op amps in the front end can be swapped for LF351 (or 353 if using duals) and get much better noise performance. Not quite as good as a 5532, but those can only be used where a dual op amp was used (5534 single did not always work, because not unity gain compensated).
In many more modern PA amps with op amps, the noise performance is limited by the front end where all the limiter circuitry is, not the power section itself. They have changed the internal gain structure quite a bit since the CS800 - power section typically runs at lower closed loop gain. That will quiet it even more.
In many more modern PA amps with op amps, the noise performance is limited by the front end where all the limiter circuitry is, not the power section itself. They have changed the internal gain structure quite a bit since the CS800 - power section typically runs at lower closed loop gain. That will quiet it even more.
I see now that I mis-interpreted the noise spec. I'm still new to the myriad "standards" used to quote S/N. Given the good THD figures, even at relatively high power output, that would indicate that the S/(N+D) is dominated by the high noise floor, no?
Excellent details - thank you. Swapping the front-end op amp is something I was eyeing as an upgrade within my skill level. As is replacing any problematic capacitors. That 2.2 uF input cap sounds like a good candidate. I've done relatively little work with op amps in my day-job, so I'll need to brush up on them, but compatibility will be primarily a function of form factor and pinout, correct?
The fan on the x runs all the time, but at low speed is not especially noticeable. Nonetheless, I may look into rigging a quieter fan in place of the current one. I've found that the fans catering to the high-end PC crowd move a surprising amount of air with a minimum amount of noise. Finding a 120V one or rigging a 12V/24V DC supply inside of the case would be the problem to solve.
That's a helpful explanation. I'm going to attempt a replacement for the op amp in the front end to see if that drops the noise floor enough for me.
The NJM4560 and the NE5532 are pin compatible. There is little wrong with the 4560, although TI substitutes for New Japan Radio may have looser tolerances. Mouser has NJM last time I checked. I found equivalent no-hiss with ST33078 on the input board instead of NJM4560, so I put the 4560 back. NJM2068 has equivalent noise spec without the wild slew rate of the next choice.
National/TI LM4562 is pin compatible and an ultimate op amp, but has such a high slew rate it may have a tendency to oscillate without special feedback resistor bypass caps. NJM4562 is an older @ inferior part, not to be bothered with.
I suggest you test or replace the input cap before mods. If it is dried up, you may have to run higher gains than necessary to hear a 1.5 vac input. Which boosts the noise.
Good luck on finding a polyprophylene cap that would fit the input board. Mouser/newark/digikey don't stock usually any under 63 v rating, and those tend to be 2" long.
Reason I refuse to buy a 2.2 uf NP electrolytic replacement cap, they come without a service life rating. I'm not putting a 500 hour red gum rubber sealed cap in something I dont want to open again. the 2.2 uf NP caps peavey used have more problems than the dozens of 10 uf or 22 uf polar caps they bought. All the polar replacement e-caps I buy are rated >3000 hours service life (10000 uf) up to 12000 hours for small ones. I had to replace TV shop e-caps in the ST70 4 times in my life, that that is too much work.
Note the CS800x is NOT the CS800 rev A,B, or C. It does not have a 741 op amp. That was the only op amp available in 1983. The CS800x is late nineties. That .02% HD figure doesn't come out of the air, it was engineeried into it. However, HD is tested at full watts. In my music room I run my amps 1/2 w to 70 watts on the SP2-XT speakers. 1/2 watts for pp orchestra, 70 watts for the cannon shot in 1812 Overture off a CD. Strumma strumma singer+guitar source tracks may never get above 2 watts.
Last edited:
I haven't used this amp, so these are just thoughts. The op amps are TL072, so there's room for improvement there, maybe NE5532 would work without any fuss. The power supply capacitors are listed as 10,000µF in the parts list, but shown as 15,000µF on the overlay, Either way that's a very small value for the size of the amplifier. I'd be putting 47,000µFs in there, or at least the largest that can be fitted. All the electros are due for replacement. Any under a few µF can be changed to polyprop. If the amp has potential any larger coupling caps with little or no voltage across them can be replaced Nichicon Muse bipolar, the rest replaced with Elna Silmic II, Cerafine or Tonerex, or Nichicon Fine Gold or other known good audio capacitors. The fan can be made quieter at low speed by increasing R103 from 400Ω to 20W to say 470Ω or 560Ω - it's a matter of trial and error. Warning: high voltages around this part!
The LF353 (or the 412 I use) is a little better than TL072. Since your amp uses duals, you can use the 5532 or 4560. You’re not stuck trying to find an appropriate single. They do draw a smidge more current, though.
I have unsuccessfully tried things like the OPA2604. On paper it’s a direct upgrade of LF353. Sounds fantastic in that position - while it works. But they do NOT tolerate the input currents that the 353 family or 5532 derivatives will when the inputs are driven beyond the rails. The input and feedback resistors won’t protect it, and as soon as the amp clips, it’s bye bye op amp. Amp stuck to the rail putting out DC to the speaker, non-recoverable except by changing the op amp. Be careful when trying really high performance op amps, as the “better” the input stage is, the more likely you can overload and damage it. Unless you add clamps, which can introduce problems of their own.
The 10,000 uF is a little small, but not by as much as you would think. That value is actually appropriate, in a hi-if sense, for stereo 8 ohm loading. For 4 ohms, all you *really* need to get full rated power down to 20 Hz is to double it. That’s where I always stop, at 22,000 uF per rail or whatever effectively gets you close. For full rated power at 20 Hz at 2 ohms of course you would need 40,000. But that might be a bit much for the rectifiers, and will increase the line current draw under all loading conditions. Anything more is useless overkill.
I have unsuccessfully tried things like the OPA2604. On paper it’s a direct upgrade of LF353. Sounds fantastic in that position - while it works. But they do NOT tolerate the input currents that the 353 family or 5532 derivatives will when the inputs are driven beyond the rails. The input and feedback resistors won’t protect it, and as soon as the amp clips, it’s bye bye op amp. Amp stuck to the rail putting out DC to the speaker, non-recoverable except by changing the op amp. Be careful when trying really high performance op amps, as the “better” the input stage is, the more likely you can overload and damage it. Unless you add clamps, which can introduce problems of their own.
The 10,000 uF is a little small, but not by as much as you would think. That value is actually appropriate, in a hi-if sense, for stereo 8 ohm loading. For 4 ohms, all you *really* need to get full rated power down to 20 Hz is to double it. That’s where I always stop, at 22,000 uF per rail or whatever effectively gets you close. For full rated power at 20 Hz at 2 ohms of course you would need 40,000. But that might be a bit much for the rectifiers, and will increase the line current draw under all loading conditions. Anything more is useless overkill.
The cs800x & cs800s have never been TL072. I've had two CS800s (due to a theft), they were NJM4560.
Some repairmen find TL071 in CS800 rev A,B,C, which would be a hiss upgrade over uA741 they were built with. But Peavey didn't install them.
Peavey did use TL074 (quad) in bargain amps like the M-2600, which has a .1% HD spec as a result.
Last edited:
That's good to know. as I said I haven't played with this amp, but the service manual for the CS800x shows TL072s.
The oversizing of the power supply capacitors has nothing to do with meeting power bandwidth specifications, it has everything to do with how the amplifiers sounds. Some people may claim that that statement is voodoo, but if they haven't been subject of a double blind test themselves, how would they know? Expectation effect cuts both ways, which means if you don't believe it you won't hear it.
Back to the amplifier in question, I can't say whether its other inherent design limitations will mask an increase in the size of the power supply capacitance, however I wouldn't even bother to set up a blind test. My past experience with evaluating sound quality of analogue amplifiers leads me to increase power supply capacitance to practical limits in any amplifier I upgrade. BTW doing so does not increase supply current - that would contravene the first law of thermodynamics AKA the law of conservation of energy. It will increase current during the first few cycles at turn on, but this amplifier has a soft-start circuit anyway.
The RMS current in the transformer does go up with increased supply capacitance. If you measure it with a typical DMM, you will not see the increase, because it only reports the current at the fundamental. The harmonic current does go up, because the charging pulses are shorter and higher in amplitude. You would only see this with a waveform analysis. Is it a problem? Sometimes.
Power bandwidth and whether it makes a difference in sound is all about power supply impedance. The impedance of the supply needs to be low compared to the load impedance. The pole frequency between the load R (parallel sum of both channels) and the supply C have the same effect as it would if it were an output capacitor. It causes a power roll at low frequency, but it is inside the feedback loop. What the feedback is doing with it determines what it’s effect has on the *sound*. Feedback’s effect isn’t always “good” by default - it is design dependent. But when the power supply impedance increases at low frequency, you will get more voltage drop in the supply under load. That causes it to clip sooner than it does above the charging frequency.
Power bandwidth and whether it makes a difference in sound is all about power supply impedance. The impedance of the supply needs to be low compared to the load impedance. The pole frequency between the load R (parallel sum of both channels) and the supply C have the same effect as it would if it were an output capacitor. It causes a power roll at low frequency, but it is inside the feedback loop. What the feedback is doing with it determines what it’s effect has on the *sound*. Feedback’s effect isn’t always “good” by default - it is design dependent. But when the power supply impedance increases at low frequency, you will get more voltage drop in the supply under load. That causes it to clip sooner than it does above the charging frequency.
Clearly power consumption is not affected by the size of the power supply capacitance. Peak currents and average currents will be different for different sized capacitors, but not the RMS current, which is the "effective" current that is used to calculate power delivered. The rms current cannot go up unless power consumption goes up because that would violate the law of conservation of energy.
It is possible that with upsized supply capacitors peak current may exceed a particular rectifier or diode rating, though I have never had an issue. Peak current rating of bridge rectifiers are typically 20-30 times the rms rating and peak currents are limited by the secondary impedance of the transformer anyway.
Larger power supply capacitors will make an amplifier sound more powerful and improve the power bandwidth to a point, but that is not the effect I was describing in the post above.
Even when an analogue amplifier is operated with plenty of headroom and is operating well within its linear operating envelope, large power supply capacitors can make it sound more effortless and/or transparent. This phenomena is not explained by power bandwidth requirements, and seems to have no upper limit, i.e. each time the supply capacitance is doubled, the effect seems to double.
It is possible that with upsized supply capacitors peak current may exceed a particular rectifier or diode rating, though I have never had an issue. Peak current rating of bridge rectifiers are typically 20-30 times the rms rating and peak currents are limited by the secondary impedance of the transformer anyway.
Larger power supply capacitors will make an amplifier sound more powerful and improve the power bandwidth to a point, but that is not the effect I was describing in the post above.
Even when an analogue amplifier is operated with plenty of headroom and is operating well within its linear operating envelope, large power supply capacitors can make it sound more effortless and/or transparent. This phenomena is not explained by power bandwidth requirements, and seems to have no upper limit, i.e. each time the supply capacitance is doubled, the effect seems to double.
Last edited:
The RMS current can indeed go up without the delivered power increasing. That was my point, and the problem. The conduction angle determines the power factor, and the bigger the cap the smaller the conduction angle and the lower the power factor.
Any effect you hear (if it’s real) of increased capacitance will be directly as a result of lower power supply impedance. Simulations always assume a perfect zero impedance voltage source and the real world can be quite different, upsetting what the amplifier is “supposed to” do.. If the amplifier becomes “better behaved” with a lower impedance power supply, you will hear improvement with bigger caps and perhaps with bypass caps that have good high frequency performance - keeping the impedance low all the way to RF. Some amps will benefit, others are so limited by other factors as to not notice a change. Some are just insensitive in that regard.
Any effect you hear (if it’s real) of increased capacitance will be directly as a result of lower power supply impedance. Simulations always assume a perfect zero impedance voltage source and the real world can be quite different, upsetting what the amplifier is “supposed to” do.. If the amplifier becomes “better behaved” with a lower impedance power supply, you will hear improvement with bigger caps and perhaps with bypass caps that have good high frequency performance - keeping the impedance low all the way to RF. Some amps will benefit, others are so limited by other factors as to not notice a change. Some are just insensitive in that regard.
Are you are confusing peak current with RMS current? By definition the RMS value of the current is the value that determines the power delivered. The Irms flowing into a power supply smoothing capacitor is by definition the same as the Irms flowing out of the capacitor, AKA the load current. The value of the load current is not affected by the size of the capacitor.
Upgrading the mains caps is silly. The OP wants to use this in his living room as a home amp. Probably with inefficient speakers he will never draw more than 100 W/ch. He could downgrade the 10000 uf rail caps to 4700 and remove 2 pairs of output transistors per channel to achieve that. The old 10000 uf caps with high ESR could achieve that current easily.
These amps are inexpensive on the used market, and provide great sound. I can't believe what e-bay auctions draw for NAD, with all the problems old ones have on here. Even the "worst SS amp made" the dynaco ST120 or ST80 is drawing $500. I bought my current cs800s, working, for $240 including freight & sales tax. The stolen CS800s with burnt out tenth watt input resistors was $100 with a whole bar band setup of ugly, dodgy, tobacco ash drenched equipment.
The mixing up in people's minds of cs800x & s with the hissy PA only 1983 CS800 A & B originals destroys the resale value of the former. Half the posters on here were offering tips to upgrade a 1983 A model. May as well run a Ford Model A at the Indianapolis Speedway.
Tip to bargain hunters. If a ebay amp is a cs800 but doesn't say x on the front, is 8.5" high, has the wind tunnel grill in the front (4"x4") and no volume knobs on the front, it is an unlabeled cs800x. There were a lot of those sold. Watch the shipping costs, model x weight >70 lb and won't ship USPS. UPS & Ex charge for distance. Why I bought a s model; it will ship USPS for $30.
These amps are inexpensive on the used market, and provide great sound. I can't believe what e-bay auctions draw for NAD, with all the problems old ones have on here. Even the "worst SS amp made" the dynaco ST120 or ST80 is drawing $500. I bought my current cs800s, working, for $240 including freight & sales tax. The stolen CS800s with burnt out tenth watt input resistors was $100 with a whole bar band setup of ugly, dodgy, tobacco ash drenched equipment.
The mixing up in people's minds of cs800x & s with the hissy PA only 1983 CS800 A & B originals destroys the resale value of the former. Half the posters on here were offering tips to upgrade a 1983 A model. May as well run a Ford Model A at the Indianapolis Speedway.
Tip to bargain hunters. If a ebay amp is a cs800 but doesn't say x on the front, is 8.5" high, has the wind tunnel grill in the front (4"x4") and no volume knobs on the front, it is an unlabeled cs800x. There were a lot of those sold. Watch the shipping costs, model x weight >70 lb and won't ship USPS. UPS & Ex charge for distance. Why I bought a s model; it will ship USPS for $30.
Last edited: