I swear....nothing is ever easy. I've been doing some cleanup around wiring and in the process changed the hookup for the frontend PSU routing the leads from the PSU to a terminal block mounted on an aluminum plate between the PSU board and the frontend board. From there I added the 100uf decoupling caps and routed the leads to each half on the frontend circuit. The Neg rail is fine but the pos rail oscillates as soon as I attach both sides. 1 side ok....2 sides about a +/-1v oscillation. added another 100uf cap at each +v connection on the board itself....no change. Have .47uf film caps mounted on the board itself. Issue with lead dressing..the PSU POs reg circuit itself? back to trial and error. didn't have this problem before I started.
Last edited:
Not sure anyone is following this thread any more. It's been a couple weeks since I posted. Still struggling trying to hunt down that darned buzz. Have replaced the Bridges, the caps (56k uf - big improvement to the sound but still buzzing) and messed around with a couple ground schemes. This amp has never sounded so good. Just wish I could get rid of the buzz. The trafo is buzzing so pulled it just to see if I could see any problems. Besides the neoprene insulators being dry and brittle didn't notice any problem. I'm thinking those need to be replaced. The amp doesn't seem to draw excessive current so I'm thinking the trafo itself is fine. The next ...and final..experiment is to replace the silpads on the output Qs. Someone, don't recall who, mentioned 15 years is a long time on a Class A amp. All that heat can cause these to fail. Anyway...ordered the boards this week from Peter to do a AJ. Time for something new. If anyone has any thoughts on the buzz problem would appreciate hearing from you. Thanks again everyone for all the help and hand holding with this repair project.
Dan
Dan
Dan-
Trafos buzz when they are being used too close to their VA rating, or if there is DC on the power line, (or perhaps in the ground). I think the buzz is made worse if you have both of the above.
I had, actually still have, an amp that buzzed (the trafo, not in the circuit) whenever my wife turned on her hairdryer upstairs. Something to do with the electrical wiring in the house. Since we moved to a new house, the amp has been perfectly silent.
Did you turn up the bias current in your re-work of the amp? Or perhaps there is something else that is causing the trafo to see high current draw?
I have solved the transformer mechanical buzz problem on two occasions by simply getting a bigger trafo, with a VA rating about 4x of the nominal usage of the amp.
JJ
Trafos buzz when they are being used too close to their VA rating, or if there is DC on the power line, (or perhaps in the ground). I think the buzz is made worse if you have both of the above.
I had, actually still have, an amp that buzzed (the trafo, not in the circuit) whenever my wife turned on her hairdryer upstairs. Something to do with the electrical wiring in the house. Since we moved to a new house, the amp has been perfectly silent.
Did you turn up the bias current in your re-work of the amp? Or perhaps there is something else that is causing the trafo to see high current draw?
I have solved the transformer mechanical buzz problem on two occasions by simply getting a bigger trafo, with a VA rating about 4x of the nominal usage of the amp.
JJ
I did re-bias the amp. I'm using 6 each of the IRFP140/9140s per side with .47R on the sources. By my calculations with 130mv across each one that gives me Class A up to about 50-60W. The trafo is a Avel torroid 60V CT rated at 550VA, the same as what was speced by NP and Norm. After reading what everyone's been using in amps this size it seems kinda low. I ran though the PS tests a month ago to measure the current draw of the amp and everything seemed to be within tolerance. Results are in Post 57.
On the AC wiring I just yesterday replaced the outlet with a hospital grade outlet. The outlet tests fine with an outlet tester but I did notice there's no separate ground wire. The box and metal conduit are acting as the ground line back to the box. wouldn't be my preferred way for an audio outlet and since this is in the basement the conduit from this box crosses alot of other conduit and water pipes on its path to the main breaker box. I'm not too comfortable working inside the mains but what I'd like to do is wire a dedicated line to the box for just my audio gear...relatively easy to do since my basement is only partially finished with no ceiling. As I'm writing you have me thinking about the AC ...again. Why? I noticed a couple weeks back that the trafo in my DAC also has a mechanical buzz. I ran a big extension cord to an upstairs out a while back with no change so discounted AC as the issue but maybe should revisit again.
Hmmmmm...... Can I use the MM to measure DC on the AC at the outlet?
Thanks JJ for the ideas.
Dan
On the AC wiring I just yesterday replaced the outlet with a hospital grade outlet. The outlet tests fine with an outlet tester but I did notice there's no separate ground wire. The box and metal conduit are acting as the ground line back to the box. wouldn't be my preferred way for an audio outlet and since this is in the basement the conduit from this box crosses alot of other conduit and water pipes on its path to the main breaker box. I'm not too comfortable working inside the mains but what I'd like to do is wire a dedicated line to the box for just my audio gear...relatively easy to do since my basement is only partially finished with no ceiling. As I'm writing you have me thinking about the AC ...again. Why? I noticed a couple weeks back that the trafo in my DAC also has a mechanical buzz. I ran a big extension cord to an upstairs out a while back with no change so discounted AC as the issue but maybe should revisit again.
Hmmmmm...... Can I use the MM to measure DC on the AC at the outlet?
Thanks JJ for the ideas.
Dan
hmm...
What you describe reminds me of my old house. A 250 VA trafo in my mini aleph buzzed, I had to switch it for a 500 VA trafo.
I think 550 VA is too small for an A-75 if you have certain issues with the wiring in your house. I wish I could say for certain whether the problem was poor grounding or DC on the mains, but I can't. I do know for certain that with better wiring those problems would go away. Many people who have not experienced it would say a trafo with a VA rating of 4x over the nominal usage is nuts.
There are some threads about DC blocking circuits on this (Pass Labs) forum, I think. One was probably started by a member who went by Choky. I remember he was one person who posted a schematic. I think there are several others though. Those threads might give you some insight as to what is causing the problem.
JJ
What you describe reminds me of my old house. A 250 VA trafo in my mini aleph buzzed, I had to switch it for a 500 VA trafo.
I think 550 VA is too small for an A-75 if you have certain issues with the wiring in your house. I wish I could say for certain whether the problem was poor grounding or DC on the mains, but I can't. I do know for certain that with better wiring those problems would go away. Many people who have not experienced it would say a trafo with a VA rating of 4x over the nominal usage is nuts.
There are some threads about DC blocking circuits on this (Pass Labs) forum, I think. One was probably started by a member who went by Choky. I remember he was one person who posted a schematic. I think there are several others though. Those threads might give you some insight as to what is causing the problem.
JJ
I came across a builder's comment somewhere here on the forum the other day describing what he did to solve an amp buzz problem experimenting with connecting directly to the chassis, through a thermistor, and through a R/D ground loop breaker circuit. I went back and tried these and using the D/R connection on the PS seems to have cleaned things up pretty well, at least while the amps is still cold. Not sure why since it had nothing to do with upstream component ground loops. Disconnected it still buzzed.
And there's almost no trafo buzz now. However, after things warms up the buzz did reappear...at least it did in my short 1 time listening test. Funny thing is it sounds alot different as compared to grounding it direct to chassis at the -IN RCA star point. If I had to describe the change I'd say it's more open and airy versus dynamic, authoratative, and forward soundstage. I think I'm heading in the right direction. If the buzz comes back I'll have to try messing around with the circuit/chassis ground points and methods. And maybe revisit the output as a possible cause if buzz remains a function of heat. As frustrating as this exercise has been I would have to say that all-in-all it has been a great learning experience. I'm with you on the trafo being a bit undersized. I'm about to start an AJ build so will probably put off upgrading the trafo for now unless I come across a bargain.
Dan
And there's almost no trafo buzz now. However, after things warms up the buzz did reappear...at least it did in my short 1 time listening test. Funny thing is it sounds alot different as compared to grounding it direct to chassis at the -IN RCA star point. If I had to describe the change I'd say it's more open and airy versus dynamic, authoratative, and forward soundstage. I think I'm heading in the right direction. If the buzz comes back I'll have to try messing around with the circuit/chassis ground points and methods. And maybe revisit the output as a possible cause if buzz remains a function of heat. As frustrating as this exercise has been I would have to say that all-in-all it has been a great learning experience. I'm with you on the trafo being a bit undersized. I'm about to start an AJ build so will probably put off upgrading the trafo for now unless I come across a bargain.
Dan
I set the bias and DC offset over the course of a week and some long listening sessions so I'm fairly confident nothing has drifted but I'll check it again. It's not easy to get at the Source resistors on the output Qs so I didn't measure the V across all of them. I guess I should rig up a probe that'll allow easier access and measure them all for peace of mind. Haven't noticed the trafo buzz volume changing much as the amp warms up since changing the ground connection to chassis via the R/D network but haven't listened for it either. Have been more focused on the speaker buzz now but will listen to the trafo again when it's fully warmed up. With the new 56k caps the AC ripple there is now only .08mv so I'm confident that's not causing it.
I'll figure this out yet. And then move on to the AJ project. Looking forward to something new. I just wish I could squeeze a little more than 25W from the AJ. With my biamped setup I know the A75 is probably overkill but not sure 25w is enough. I like my music LOUD.....as in concert level. Must be the rock'n rolla character flaw. Dan
I'll figure this out yet. And then move on to the AJ project. Looking forward to something new. I just wish I could squeeze a little more than 25W from the AJ. With my biamped setup I know the A75 is probably overkill but not sure 25w is enough. I like my music LOUD.....as in concert level. Must be the rock'n rolla character flaw. Dan
Fixed.
Rookie mistake. I changed the location of the output grounds to the starpoint between the main caps. Previously I had it connected to the same point as the trafo CT. Even though it's only a 3-4 inches from where it was. It's absolutely dead quiet now. I also went ahead and measured the bias across each output R. Across the 12 right side Qs the power output varies from a high of 11.3W to a low of 8.9W. The other side is 10.7-7.0W. Not bad. Also wondering about dropping the output Q count down from 12 to 6, which would double the W each Q would have to sink (20-25W). It seems like most amps this size are only using half as many output Qs as the A75. Do you think this would improve the sound? The IRFP140/9140s are rated at 180W each so they are certainly capable of handling the extra power.
Feels good when a plan finally comes together.
Rookie mistake. I changed the location of the output grounds to the starpoint between the main caps. Previously I had it connected to the same point as the trafo CT. Even though it's only a 3-4 inches from where it was. It's absolutely dead quiet now. I also went ahead and measured the bias across each output R. Across the 12 right side Qs the power output varies from a high of 11.3W to a low of 8.9W. The other side is 10.7-7.0W. Not bad. Also wondering about dropping the output Q count down from 12 to 6, which would double the W each Q would have to sink (20-25W). It seems like most amps this size are only using half as many output Qs as the A75. Do you think this would improve the sound? The IRFP140/9140s are rated at 180W each so they are certainly capable of handling the extra power. Feels good when a plan finally comes together.
Dan - Glad you've got it quiet. Makes me want to work on mine to get that last little bit of hum out.
Kannan - The benefit of fewer output devices is decreased total gate capacitance, which makes the output stage easier to drive at high frequencies. This may explain some of the sonic differences between the Aleph variants. Whether it's audible with the A75 and it's higher current VAS is subject to testing. I like my 6 pair amp, but haven't tried more devices.
Kannan - The benefit of fewer output devices is decreased total gate capacitance, which makes the output stage easier to drive at high frequencies. This may explain some of the sonic differences between the Aleph variants. Whether it's audible with the A75 and it's higher current VAS is subject to testing. I like my 6 pair amp, but haven't tried more devices.
My point is about increasing the current to compensate the Class A loss due to lowered output Transistors. yes he can try lowering the outputs and see whether he can hear the differences also try increasing the current slowly. Dan has worked with Class A bias anyway.
He is now having the courage to test new things on his A75. I Wish him luck !
have been reading his
He is now having the courage to test new things on his A75. I Wish him luck !
have been reading his
It sounds pretty good. Better than before but I would attribute that to the bigger caps. I upgraded those from 20k to 56k. I still think it sounded best when I had the ground / chassis connection right at the -IN between the RCAs but the buzz was alot louder then. I might still try that ground point again through a Thermister or R/D. Right now it's grounded through the RD on the PS board with a 14GA wire directly to the power entry ground. I've always used the 2-stage gain configuration with no feedback. I find that feedback around the output is a litttle too agresive. No feedback is more natural and the soundstage is more open.
So you don't think there's any benefit soundwise to use fewer outputs at a higher bias? I've noticed most of the Aleph and F series amp go with far fewer and was thinking there might be some benefit going that route with the A75 as well.
So you don't think there's any benefit soundwise to use fewer outputs at a higher bias? I've noticed most of the Aleph and F series amp go with far fewer and was thinking there might be some benefit going that route with the A75 as well.
It's definitely been an experience. Educational and frustrating. Part of the exercise has been focused on circuit performance and the other on listening. I actually played music professionally for 10 years back in the mid-70-80s and always reference the "live" sound to my music system. Right now I'd say it's very close. My distaste with transistor is it tends to sound too etched and aggressive in the mids and upper end. Being a guitar player I guess would explain that. When it comes to my guitar amps it's either tubes or nothing. I like smooth and liquid with a wide & deep soundstage. I'd say right now with the way it's set up it's about 90% there. The rest of the system consists of Cabbasse 6"mids, Accuton ceramic domes, some NHT12 inch subs. My preamp is also a biamp using a LR filter crossed at about 100hz. I drive the subs with a big Carver. The CD transport is an old Pink Triangle Cardinal which feeds my DAC with a BurrBrown PCM63 via BNC.
Now that the A75 is up and running it's time to start the AJ project. It's looks less complicated than the A75 and should be a fun build. It's been many years since I built anything. Not sure I can manage with only 25w. I'd really like at least 50w/channel.
Dan
Now that the A75 is up and running it's time to start the AJ project. It's looks less complicated than the A75 and should be a fun build. It's been many years since I built anything. Not sure I can manage with only 25w. I'd really like at least 50w/channel.
Dan
Bob,
If you're running just 6 pairs that's 25W each. Are you using fans or had any problems with case temperatures? Or did I misunderstand...6 Ps and 6 Ns. For what's it's worth I have plenty of top end running the with 12 Qs (6P/6N). My ears aren't that sensative to highs anymore so not sure I could tell the difference if it had just 6.
Dan
BTW...thanks for all the help with this project.
If you're running just 6 pairs that's 25W each. Are you using fans or had any problems with case temperatures? Or did I misunderstand...6 Ps and 6 Ns. For what's it's worth I have plenty of top end running the with 12 Qs (6P/6N). My ears aren't that sensative to highs anymore so not sure I could tell the difference if it had just 6.
Dan
BTW...thanks for all the help with this project.
With fewer devices you could lower the bias a bit to keep the devices in their safe operating area at the cost of lower output in class A, but the amp will still operate quite nicely in class AB to over 75W. As long as you have enough heat sink you can run 25W per device. Just calculate the junction temperature and try to keep it under 100C.
My A75s are not stock and were a bit of a busy first build. I figured that I needed more than 75W because my 75 WPC receiver sounded fuzzy on loud passages. I scaled it up a bit somewhat based on what I could find for surplus power supply parts and ended up with 57V output rails and 65V front end, tweaking some part values and voltage or power ratings to get there. I am at about 1.75A total bias per channel, which should give me about 50W class A and somewhere north of 150W class AB if needed.
That works out to ~16W per device. Heat sinks are 10" vertical 14" deep with 32 1.5" fins. They get to around 50C using NP's touch temperature scale.
Were I to do it again, I would probably build a standard A75, as I rarely exceed 20W peaks.
My A75s are not stock and were a bit of a busy first build. I figured that I needed more than 75W because my 75 WPC receiver sounded fuzzy on loud passages. I scaled it up a bit somewhat based on what I could find for surplus power supply parts and ended up with 57V output rails and 65V front end, tweaking some part values and voltage or power ratings to get there. I am at about 1.75A total bias per channel, which should give me about 50W class A and somewhere north of 150W class AB if needed.
That works out to ~16W per device. Heat sinks are 10" vertical 14" deep with 32 1.5" fins. They get to around 50C using NP's touch temperature scale.
Were I to do it again, I would probably build a standard A75, as I rarely exceed 20W peaks.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.