Progress
I've gotten my RevA running again today. I needed to improve the fit of the chips on the heatsinks.
I found they needed what I consider excessive gain to drive them.
With max level from my pre-amp they were at normal listening levels. This drive level was above what I normally use with a 35 w/ch stereo reciever and 6db boost added through a Marchand active x-over. I took the x-over out of the loop and got the same results. Taking the x-over out also left a low-level hum in the system.
So, I'm going to check my resistor values, starting with the gain stages. I don't know where to start with chasing the hum away. We'll try one thing at a time. Suggestions are welcome.
I've gotten my RevA running again today. I needed to improve the fit of the chips on the heatsinks.
I found they needed what I consider excessive gain to drive them.
With max level from my pre-amp they were at normal listening levels. This drive level was above what I normally use with a 35 w/ch stereo reciever and 6db boost added through a Marchand active x-over. I took the x-over out of the loop and got the same results. Taking the x-over out also left a low-level hum in the system.
So, I'm going to check my resistor values, starting with the gain stages. I don't know where to start with chasing the hum away. We'll try one thing at a time. Suggestions are welcome.
maf_au said:
That will work fine. As long as red and orange are connected to the ac inputs, the wire and black is not critical. On this circuit anyway.
I see what those cost. Even corrected for Aussie dollars to UDS, they are still not cheap. For US, I had some custom made in Atlanta at Victoria Magnetics that are 60 - 65.00 for a 300 watt, toriod. These have electrostatic shields to boot. I think they are better than the hallowed Plitron.
But bang for the buck, the 48 volt centertapped ones from PE, are unbeatable. Some experts feel that E core is better than toriod. The bandwidth of the toriod lets more mains garbage through.
Your amp is looking good!
George
Re: Progress
Very curious, as the gain on my REV A actually seems on the high side. I actually have to go with lower volumes compared to my "Inverting NOT" amp...
If you are getting hum it is most likely coming from you mains GND. Just as a test try disconnecting it from the REV A.
I am sorry I am not more help... Hard to concentrate when 2-3lb small mouth bass are waiting outside your back porch.
Back to the creek...
I will be thinking it over though...
Cheers!
Russ
bg40403 said:
Very curious, as the gain on my REV A actually seems on the high side. I actually have to go with lower volumes compared to my "Inverting NOT" amp...
If you are getting hum it is most likely coming from you mains GND. Just as a test try disconnecting it from the REV A.
I am sorry I am not more help... Hard to concentrate when 2-3lb small mouth bass are waiting outside your back porch.
Back to the creek...
I will be thinking it over though...
Cheers!
Russ
Re: Progress
Since it works, it should be easy to find. Check ac input, should be 24 - 25 volts. Then DC on each channel, 34 - 35 volts on each rail.
Check dc offset for each channel, mine is 10-15 mv, some report lower. If it is mv level you are fine.
Since the LM3886 is unity gain, the LM318 provides all the gain. The ratio of the 100k input resistor is important, along with the 12K feedback resistor.
Also recheck what you have for small film caps. I almost strapped a 100 nf cap around the 318 instead of the required 100 pf. Think this would have created some issues.
The hum makes me think there is a grounding issue. This may be the gain issue also. Check ground from the body of the RCA to the power supply ground. Should be 1 ohm. Check the RCA from center to body, should read 100K. Check RCA center pin to input film cap, should read 3.3K If these all check the input section is right, try looking over the feedback loop. Maybe polarity of the electrolytic.
George
bg40403 said:
Since it works, it should be easy to find. Check ac input, should be 24 - 25 volts. Then DC on each channel, 34 - 35 volts on each rail.
Check dc offset for each channel, mine is 10-15 mv, some report lower. If it is mv level you are fine.
Since the LM3886 is unity gain, the LM318 provides all the gain. The ratio of the 100k input resistor is important, along with the 12K feedback resistor.
Also recheck what you have for small film caps. I almost strapped a 100 nf cap around the 318 instead of the required 100 pf. Think this would have created some issues.
The hum makes me think there is a grounding issue. This may be the gain issue also. Check ground from the body of the RCA to the power supply ground. Should be 1 ohm. Check the RCA from center to body, should read 100K. Check RCA center pin to input film cap, should read 3.3K If these all check the input section is right, try looking over the feedback loop. Maybe polarity of the electrolytic.
George
Panelhead, Thank You
I started by checking for ground issues. Right off the bat I found something puzzling about my input resistance readings. I crossed my input leads! I was confused by what I thought (I ***umed) was a "minus" sign at the input...Nope, looks like the numeral "1" now...Oh well.
I'm switching the leads and will do some checks, then back to the speakers.
I started by checking for ground issues. Right off the bat I found something puzzling about my input resistance readings. I crossed my input leads! I was confused by what I thought (I ***umed) was a "minus" sign at the input...Nope, looks like the numeral "1" now...Oh well.
I'm switching the leads and will do some checks, then back to the speakers.
Russ, George:
Thanks for the help and encouragement, if it sounds as good as it looks I'll be happy
There seems to be no such thing as a good, cheap transformer here in Australia. Shipping them from the US would add a lot of freight, and risk of damage because of their weight.
Unfortunately, I'm running out of time to finish it today, gotta hop on a plane in a few hours... Looks like I'll have to finish it Friday night.
Michael
Thanks for the help and encouragement, if it sounds as good as it looks I'll be happy
There seems to be no such thing as a good, cheap transformer here in Australia. Shipping them from the US would add a lot of freight, and risk of damage because of their weight.
Unfortunately, I'm running out of time to finish it today, gotta hop on a plane in a few hours... Looks like I'll have to finish it Friday night.
Michael
Russ,
With the monoblocks, is the pcb sized such that a pair of them can be attached to a heatsink that has holes drilled to suit the My_REF?
If not, please let me know what the minimum hole-to-hole spacing is and I'll predrill extra holes to suit, I'm sure I have adequate width...
I'm thinking it would make testing them quick if I could swap them into the same case. I've got 12cm from heatsink to case divider.
Michael
With the monoblocks, is the pcb sized such that a pair of them can be attached to a heatsink that has holes drilled to suit the My_REF?
If not, please let me know what the minimum hole-to-hole spacing is and I'll predrill extra holes to suit, I'm sure I have adequate width...
I'm thinking it would make testing them quick if I could swap them into the same case. I've got 12cm from heatsink to case divider.
Michael
Go figure
Switching the input leads did solve the gain problem.
The hum remains.
Some measurements: R/L
ACV secondary 25.2/25.17
DCV rails 32.4/32.4
DC offset mV 6/7
I need to get better access to the input to measure and inspect solder joints. Some disassembly comes next.
RCA body to ground, ohms 0.4/1.4
RCA pin to body NR/NR
Those are suspect and I'll also check across the input cap. I had the hum with either input connected. It is on both channels.
Switching the input leads did solve the gain problem.
The hum remains.
Some measurements: R/L
ACV secondary 25.2/25.17
DCV rails 32.4/32.4
DC offset mV 6/7
I need to get better access to the input to measure and inspect solder joints. Some disassembly comes next.
RCA body to ground, ohms 0.4/1.4
RCA pin to body NR/NR
Those are suspect and I'll also check across the input cap. I had the hum with either input connected. It is on both channels.
HI Ed,
Your power measuerments are nearly identical to mine, including offset.
What, is odd is that my resistance from RCA body to GND is exactly 1ohm. I suspect your mains GND and pwer GND are connected. Am I correct? Either at the RCA or some other place. If so that could be your hum.
Cheers!
Russ
Your power measuerments are nearly identical to mine, including offset.
What, is odd is that my resistance from RCA body to GND is exactly 1ohm. I suspect your mains GND and pwer GND are connected. Am I correct? Either at the RCA or some other place. If so that could be your hum.
Cheers!
Russ
Progress
I found that I had the output grounds swapped. Again, I assumed they would remain on their respective sides which was not the case. (Picture the fun I've had on Christmas' past when putting toys together at the 11th hour. Instructions? I don't need no freakin' instructions!.....How many times have I paid for that?....)
This is the reason the RCA body to ground measurement didn't check. Now I measure 1.2 ohms on both. This after putting fresh batteries in my DVOM.
I still get no reading for resistance between the RCA pin to body. I'll make some shorting jacks tomorrow and continue the search. In spite of the hum, which is less now, I can hear the quality of this amp will be worth the effort.
I found that I had the output grounds swapped. Again, I assumed they would remain on their respective sides which was not the case. (Picture the fun I've had on Christmas' past when putting toys together at the 11th hour. Instructions? I don't need no freakin' instructions!.....How many times have I paid for that?....)
This is the reason the RCA body to ground measurement didn't check. Now I measure 1.2 ohms on both. This after putting fresh batteries in my DVOM.
I still get no reading for resistance between the RCA pin to body. I'll make some shorting jacks tomorrow and continue the search. In spite of the hum, which is less now, I can hear the quality of this amp will be worth the effort.
My input AC power ground is connected to the chassis. Nothing else. I may put some bypassing in later but for now, there is nothing connecting DC power output or amplifier signal ground to anything else. Only the fastons and the input connectors straight to their off-board connectors and the off-board connectors are isolated from the chassis. I hear no hum, or much of anything except the music.
Re: Ask for guidance
Ed there is no good reason to connect you mains ground and your power ground. The main ground is just there for safety, in case the case somehow gets mains voltage to it.
I always just attach the mains ground to the chasis, but the power ground should be isolated to the power supply/circuit(in this case all on the same board).
There is absolutely no beneifit to connecting mains GND to anything in the circuit, in fact it can really only cause problems, as the circuit's GND is really a voltage reference, not a safety feature.
Once again the Mains ground is there to protect the user from mains voltage on the case. Nothing more.
Most of this I am stating for the noobs, not so much to you Ed.
The best way to keep them isolated is to use RCA jacks that have an insulator as well as the same type of binding posts. I would suspect your RCA connectors based on your symptoms.
Cheers!
Russ
bg40403 said:
Ed there is no good reason to connect you mains ground and your power ground. The main ground is just there for safety, in case the case somehow gets mains voltage to it.
I always just attach the mains ground to the chasis, but the power ground should be isolated to the power supply/circuit(in this case all on the same board).
There is absolutely no beneifit to connecting mains GND to anything in the circuit, in fact it can really only cause problems, as the circuit's GND is really a voltage reference, not a safety feature.
Once again the Mains ground is there to protect the user from mains voltage on the case. Nothing more.
Most of this I am stating for the noobs, not so much to you Ed.
The best way to keep them isolated is to use RCA jacks that have an insulator as well as the same type of binding posts. I would suspect your RCA connectors based on your symptoms.
Cheers!
Russ
Don't do as I do, do as I say
It's a necessary phrase for raising children...
Let me share with you my experiences;
The amp is running fine, no hum.
I had a combination of faults
My RCA's were not isolated from the chassis. I made the holes larger so the "barrel" of the jack didn't come in contact with the case.
I removed the...ahem...strap from amp ground to case ground...noob?...that would be me...
I checked isolation of the speaker posts from the case.
When I plugged it in through the Marchand active x-over, the hum was still there. I removed the Marchand from the loop and the hum is gone. Thank God!
Seems I've had a ground problem I wasn't aware of in the Marchand. Guess which component I'll open up next? I suppose the reciever I was using before wasn't suseptible in the same way.
Thanks to all: Russ, hayenc and Panelhead for your responses in helping me get through this.
My pre-amp output runs 0-90. I had been listening around 65-70 with a 35w/ch amp. This amp rums fine (louder) set around 55. no gain problem. I'm just beginning the use of the amp now. That's Little Feat's "Last Record Album" playing in the background. You really ought to find a copy and give it a listen.
It's a necessary phrase for raising children...
Let me share with you my experiences;
The amp is running fine, no hum.
I had a combination of faults
My RCA's were not isolated from the chassis. I made the holes larger so the "barrel" of the jack didn't come in contact with the case.
I removed the...ahem...strap from amp ground to case ground...noob?...that would be me...
I checked isolation of the speaker posts from the case.
When I plugged it in through the Marchand active x-over, the hum was still there. I removed the Marchand from the loop and the hum is gone. Thank God!
Seems I've had a ground problem I wasn't aware of in the Marchand. Guess which component I'll open up next? I suppose the reciever I was using before wasn't suseptible in the same way.
Thanks to all: Russ, hayenc and Panelhead for your responses in helping me get through this.
My pre-amp output runs 0-90. I had been listening around 65-70 with a 35w/ch amp. This amp rums fine (louder) set around 55. no gain problem. I'm just beginning the use of the amp now. That's Little Feat's "Last Record Album" playing in the background. You really ought to find a copy and give it a listen.
my findings
I followed this thread with great interest and I was a bit sceptical at first. But at the same time excited about a new approach. I was getting tired of the "micro tweaking" (add a snubber and remove a cap here and there) of the chip amp and nobody really could give a fresh approach.
I myself am a JLH fan an I have my configuration build up as monoblocks. I also have tried all the "silly" tweaks and the good ones suggested on the web. The capacitance multiplier being outright ridiculous. But I am wondering here.
Over the weekend I quickly whipped up a Rev-A. Just to have good grounds for further comparisons and Mauro, you have done a BRILLIANT job with this design.
I can answer a couple of questions that has not been addressed.
1- putting an OPA627 into the equation does not improve it. It sounds the same.
2- you need a LM317 in order to run Rev-C, it uses pin 8 with you don't on other op-amps (it is a very good and cheap little op-amp that does a good job).
I have build mine with twin bridge rectifiers (just a habit I have when building amps) this make it easier to add another transformer at a later stage, and splitting the rails properly. This also brings the noise down by a large margin, and you do not get any cross talk
I do not have the speaker protection circuit in my system. (Being experimental)
Turn-on thumps. This is the strange thing. I have no turn-on thumps, just a small click. OK then I have to admit I do not have any DC protection. But this is only an experiment and I will be doing the Rev-C next. With the protection
Listening impressions
The first things that I found is that you have a bit more accentuation on the high mids and tops. Making it sound “forward” as explained by Russ in his evaluation. I did find however that the tonality is not as resolved or rather rounded as my JLH and the Bass is not as deep (it is however much better than a standard JLH). BUT having said that. The bass is extremely good. And the overall balance is much better that anything else I have heard. If you still wander how it compares to a chip amp with all the micro tweaking. It is A LOT better by a large margin. I believe that the REV-C will be the preferred version at the end of the day specially because the tonal balance will be in line of what it should be without the accent on the high mids and tops.
Suggestions.
Split the bridge rectifiers giving you two instead of the one per channel (this way you decouple the transformer from earth). And a higher voltage input cap, they do tend to sound a lot better that the small caps (why I do not know).
Mauro I have to congratulate you on a very good design I am impressed.
Russ, keep up the good work. You will have a Gem to offer the people here
Regards
Rudi
I followed this thread with great interest and I was a bit sceptical at first. But at the same time excited about a new approach. I was getting tired of the "micro tweaking" (add a snubber and remove a cap here and there) of the chip amp and nobody really could give a fresh approach.
I myself am a JLH fan an I have my configuration build up as monoblocks. I also have tried all the "silly" tweaks and the good ones suggested on the web. The capacitance multiplier being outright ridiculous. But I am wondering here.
Over the weekend I quickly whipped up a Rev-A. Just to have good grounds for further comparisons and Mauro, you have done a BRILLIANT job with this design.
I can answer a couple of questions that has not been addressed.
1- putting an OPA627 into the equation does not improve it. It sounds the same.
2- you need a LM317 in order to run Rev-C, it uses pin 8 with you don't on other op-amps (it is a very good and cheap little op-amp that does a good job).
I have build mine with twin bridge rectifiers (just a habit I have when building amps) this make it easier to add another transformer at a later stage, and splitting the rails properly. This also brings the noise down by a large margin, and you do not get any cross talk
I do not have the speaker protection circuit in my system. (Being experimental)
Turn-on thumps. This is the strange thing. I have no turn-on thumps, just a small click. OK then I have to admit I do not have any DC protection. But this is only an experiment and I will be doing the Rev-C next. With the protection
Listening impressions
The first things that I found is that you have a bit more accentuation on the high mids and tops. Making it sound “forward” as explained by Russ in his evaluation. I did find however that the tonality is not as resolved or rather rounded as my JLH and the Bass is not as deep (it is however much better than a standard JLH). BUT having said that. The bass is extremely good. And the overall balance is much better that anything else I have heard. If you still wander how it compares to a chip amp with all the micro tweaking. It is A LOT better by a large margin. I believe that the REV-C will be the preferred version at the end of the day specially because the tonal balance will be in line of what it should be without the accent on the high mids and tops.
Suggestions.
Split the bridge rectifiers giving you two instead of the one per channel (this way you decouple the transformer from earth). And a higher voltage input cap, they do tend to sound a lot better that the small caps (why I do not know).
Mauro I have to congratulate you on a very good design I am impressed.
Russ, keep up the good work. You will have a Gem to offer the people here
Regards
Rudi
Attachments
Re: my findings
I have build mine with twin bridge rectifiers (just a habit I have when building amps) this make it easier to add another transformer at a later stage, and splitting the rails properly. This also brings the noise down by a large margin, and you do not get any cross talk
I have always used dual rectifier bnridges for dual pail supplies. A frind who builds a lot more products than I do uses a single bridge and references ground. He says the single bridge scopes better. I have used both and cannot hear a difference.
I can see, where a dual bridge should have lower value switching spikes. Just twice as many. Mauro must prefer this method.
I do not have the speaker protection circuit in my system. (Being experimental)
Turn-on thumps. This is the strange thing. I have no turn-on thumps, just a small click. OK then I have to admit I do not have any DC protection. But this is only an experiment and I will be doing the Rev-C next. With the protection
I bumped up the on board filter capacitance for the circuit. There is a no turn on transient, but a small turn off pop sometimes. It seems heat related, worse after a hard workout.
I think if it was built using the stock value filter caps mine might be quiet on turn off too. Could not bring myself to use so little capacitance for an amplifer. Usually 150,000 ufd to 250,000 ufd per channel is about right.
It is a nice circuit. I might add it to some other amps I have laying around.
What value of dc on the output wil trigger it to open the relay? I guess 0.6 - 0.7 volts.
Agree 100% on the sonics. The RevA is not perfect. But is close. I suspect the Rev C will be a little better.
Many thanks to Mauro for sharing the design, and to Russ for making this happen.
George
I have build mine with twin bridge rectifiers (just a habit I have when building amps) this make it easier to add another transformer at a later stage, and splitting the rails properly. This also brings the noise down by a large margin, and you do not get any cross talk
I have always used dual rectifier bnridges for dual pail supplies. A frind who builds a lot more products than I do uses a single bridge and references ground. He says the single bridge scopes better. I have used both and cannot hear a difference.
I can see, where a dual bridge should have lower value switching spikes. Just twice as many. Mauro must prefer this method.
I do not have the speaker protection circuit in my system. (Being experimental)
Turn-on thumps. This is the strange thing. I have no turn-on thumps, just a small click. OK then I have to admit I do not have any DC protection. But this is only an experiment and I will be doing the Rev-C next. With the protection
I bumped up the on board filter capacitance for the circuit. There is a no turn on transient, but a small turn off pop sometimes. It seems heat related, worse after a hard workout.
I think if it was built using the stock value filter caps mine might be quiet on turn off too. Could not bring myself to use so little capacitance for an amplifer. Usually 150,000 ufd to 250,000 ufd per channel is about right.
It is a nice circuit. I might add it to some other amps I have laying around.
What value of dc on the output wil trigger it to open the relay? I guess 0.6 - 0.7 volts.
Agree 100% on the sonics. The RevA is not perfect. But is close. I suspect the Rev C will be a little better.
Many thanks to Mauro for sharing the design, and to Russ for making this happen.
George