Terry I am glad to hear it.
I think you are lucky though that you didn't blow them again though. We should have been more clear on this - but for others who may read this thread - it is important that when doing this type of thing that the bias is turned all the way down first (ie max resistance) before powering the amp.
Anyway good one - now take a well deserved brake and listen to some tunes.
Arius - I wonder why it is that Leach didn't increase the value of this resistor in the first place. It has always been a question that has bugged me. I am not sure what my exact value is on the pot, however I know that it is less than half way.
I think you are lucky though that you didn't blow them again though. We should have been more clear on this - but for others who may read this thread - it is important that when doing this type of thing that the bias is turned all the way down first (ie max resistance) before powering the amp.
Anyway good one - now take a well deserved brake and listen to some tunes.
Arius - I wonder why it is that Leach didn't increase the value of this resistor in the first place. It has always been a question that has bugged me. I am not sure what my exact value is on the pot, however I know that it is less than half way.
Hi,
when first setting up a power amp it is better to short the input to ground. This is easily achieved by inserting a phono plug into the input with a shorting link soldered from pin to ground. Some retailers sell these shorting phonos but they are very easy to make.
The shorted input prevents spurious signals reaching the first amp stage and some amps can become unstable when the input is not connected to a source.
Terry,
what is the output stage Vcc when the input stage Vc=+-68Vdc and your bias is set at your chosen level? There was some advice earlier that these two voltages should be set very similar or within about 2V from memory.
when first setting up a power amp it is better to short the input to ground. This is easily achieved by inserting a phono plug into the input with a shorting link soldered from pin to ground. Some retailers sell these shorting phonos but they are very easy to make.
The shorted input prevents spurious signals reaching the first amp stage and some amps can become unstable when the input is not connected to a source.
Terry,
what is the output stage Vcc when the input stage Vc=+-68Vdc and your bias is set at your chosen level? There was some advice earlier that these two voltages should be set very similar or within about 2V from memory.
AndrewT said:
Hi Andrew,
I just took measurements while it is playing and it reads 67.4 coming out of the regulator and 64.3 on the VSS & VCC rails. When I took measurements yesterday, I was getting 1.1V at the base of the drivers which is what the backend sees and which is what Jens listed on the chart he posted so I figured I was good to go. Is there a reason that the 2V differancial is important?
Hi Arius,
It doesn't sound like smoke and mirrors to me. I tried what you suggested, except for my missunderstanding what you meant by maxing out the resistance. I explained that earilier. For some reason, removing the protection circuit on mine caused something really weird to happen but it was probably something I did wrong. I did learn a lot from this exsorsize so all is well.
Thanks so much all of you for taking the time to help me through this. It is all a learning experience for me and I feel like I have learned quite a bit and I hope others will benefit from it as well.
The amp is sounding wonderful.
I tried setting the bias using my mA meter but for some reason it wouldn't work properly so I set it to 50mV across a pair of emitter resistors and it is running at about 31c on the heatsinks. Drums sound very natural. I think I'll leave it right there for now.
Blessings, Terry
Yes, let's just leave it there for now. I'm really happy for you. I'm still working on my Leach. Trying to keep the costs down. Just ordered my case and toroidals today. Ouchie.
I am not Andrew but this is probably due to fear of clipping the output stage when the input is overdriven. It's bad for the output stage to clip, so if an amp has to clip, we'd rather it happens at the front end.
pooge said:
I did after it ran for a couple of hours. I really couldn't hear any difference and it raised the heatsink temp to 42c. I went ahead and left it there anyway. After a couple of days of listening I may try dropping it back down a little. If I don't hear any change, I can't see wasting the electricity.
Blessings, Terry
OMG - Well I just recieved my 12 boards. They are beatifully made and with 2.5 Ounce Copper.
Then I took a closer look. The production house had transposed the bottom & top copper layers
. Anyway it is not too tragic - it only realy has an effect on the thermal sensors. Th rest of the components I am beleive are mountable. Anybody disagree?
The good thing is they will be sending me a new batch of corrected boards for nothing and do not want the old ones back. So now I can build 24 channels
.
All of my active components have now arrived - just waiting on quotes from the transformer guy. I am thinking of 40-0-40 400VA. Any comments?
Thx Jens for this layout - makes things alot easier
Tony - We havn't forgotten about the protection circuit
Terry - Things still all on track?
Then I took a closer look. The production house had transposed the bottom & top copper layers
. Anyway it is not too tragic - it only realy has an effect on the thermal sensors. Th rest of the components I am beleive are mountable. Anybody disagree?The good thing is they will be sending me a new batch of corrected boards for nothing and do not want the old ones back. So now I can build 24 channels
.All of my active components have now arrived - just waiting on quotes from the transformer guy. I am thinking of 40-0-40 400VA. Any comments?
Thx Jens for this layout - makes things alot easier
Tony - We havn't forgotten about the protection circuit
Terry - Things still all on track?
Byrd said:OMG - Well I just recieved my 12 boards. They are beatifully made and with 2.5 Ounce Copper.
Then I took a closer look. The production house had transposed the bottom & top copper layers
The good thing is they will be sending me a new batch of corrected boards for nothing and do not want the old ones back. So now I can build 24 channels
All of my active components have now arrived - just waiting on quotes from the transformer guy. I am thinking of 40-0-40 400VA. Any comments?
Thx Jens for this layout - makes things alot easier
Tony - We havn't forgotten about the protection circuit
Terry - Things still all on track?
Yup, still on track. This is a very powerful amp. I didn't realize it would have so much kick. It is very sanitary sounding compared to my other amps. I remember now when someone here compared a chip amp to one of these and said how much more "information" was there when listening to the Leach. I completely agree with that statement. I think I can hear the engineer breathing from the booth.
It's a little too clear for some of the recordings I have. I don't think the mixing engineers had quite this accurate an amp to mix through.
Blessings Terry
pooge said:
Hi pooge,
Yes, I have one in my P101. It is dead quiet as far as I can tell. I also have one that I bought for my Double Barrel but I haven't got that running yet. I have however powered it up before and don't recall any mechanical noise at all. Both of them put out the voltage that they were represented to do. I can't complain at all. I don't own a working scope so I can't tell what I can't hear. It will make noise if you hook up the filter caps backward though. I can attest to that.
Blessings, Terry
Hy
I need help on leach clone...,
I vant to use 63V capacitors in psu so wath is the max. voltage that is safe to use vith 63V caps., I think to use +-52V, and regulated voltage for front end, that is 10V more then main psu, do I have to calculate missing resistor values to main PSU or to regulateed,I need help to calculate protection resistors.
Is it decoupling capacitor to big, original vlue is 220uF, some other capacitors value ( c13,c4,c11...) is oversized, I want to order some parts and i want to be sure....., can I switch input bipolar capacitor to 2,2uF epcos MKT, In Burklin katalog I finde Frolyt EKSU bipolar capacitors, is it Ok ( I never heard for that brand ).
Best regards, Vinko
I need help on leach clone...,
I vant to use 63V capacitors in psu so wath is the max. voltage that is safe to use vith 63V caps., I think to use +-52V, and regulated voltage for front end, that is 10V more then main psu, do I have to calculate missing resistor values to main PSU or to regulateed,I need help to calculate protection resistors.
Is it decoupling capacitor to big, original vlue is 220uF, some other capacitors value ( c13,c4,c11...) is oversized, I want to order some parts and i want to be sure....., can I switch input bipolar capacitor to 2,2uF epcos MKT, In Burklin katalog I finde Frolyt EKSU bipolar capacitors, is it Ok ( I never heard for that brand ).
Best regards, Vinko
Is there any way yet of calculating R68 / R69?
I have used Jaco's method of calculation and have come up with
R28/R30 270R
R41/R42/R43 and R60/R61/R62 1.5K
R26/R34 4.3k
For the MJE4* devices at 61.5 Volts (for surge).
If not what components are required to be left out when populating the board?
I have used Jaco's method of calculation and have come up with
R28/R30 270R
R41/R42/R43 and R60/R61/R62 1.5K
R26/R34 4.3k
For the MJE4* devices at 61.5 Volts (for surge).
If not what components are required to be left out when populating the board?
Hi Pinco,
you can use upto about +-57Vdc when the output is unloaded on a 63Vdc capacitor. This leaves +10.5% for mains overvoltage. I plan on using 40Vac which will give +-58Vdc on 63Vdc caps reducing my overvoltage margin, but I am willing to take that risk.
Leach Clone is a BJT output stage and does not need a voltage increase for the driver and volt amp stages, unlike FETs which do need an 8V to 10V extra drive.
The decoupling caps on Leach Clone are 10,000uF (10mF) not 220uF, but you could use anything between 4,700uF (4.7mF) and 15,000uF (15mF).
C4, C11 & C13 could be reduced to 100nF, since they are bypass caps, if you are having trouble sourcing caps to fit.
In my opinion C7 is far too big. The time constant C7*R17 should be about half an octave higher than C12*R16. Values from 2u2F to 15uF would seem better for the input cap and you can get PP in these values.
you can use upto about +-57Vdc when the output is unloaded on a 63Vdc capacitor. This leaves +10.5% for mains overvoltage. I plan on using 40Vac which will give +-58Vdc on 63Vdc caps reducing my overvoltage margin, but I am willing to take that risk.
Leach Clone is a BJT output stage and does not need a voltage increase for the driver and volt amp stages, unlike FETs which do need an 8V to 10V extra drive.
The decoupling caps on Leach Clone are 10,000uF (10mF) not 220uF, but you could use anything between 4,700uF (4.7mF) and 15,000uF (15mF).
C4, C11 & C13 could be reduced to 100nF, since they are bypass caps, if you are having trouble sourcing caps to fit.
In my opinion C7 is far too big. The time constant C7*R17 should be about half an octave higher than C12*R16. Values from 2u2F to 15uF would seem better for the input cap and you can get PP in these values.
Hi Byrd,
the protection values used by Leach and repeated on this thread use a constant current curve when output voltage is less than zero (negative output). This leads to severe overloading of the output stage before the protection cuts in when driving reactive loads.
The ideal curve would be a falling limiting current as the negative output voltage increases.
Using R68 would achieve this but it's value has to be designed as a matched set for the whole protection locus. It cannot be added to an existing set of Leach resistors.
I posted a set of values for the conditions applying to Still4given's set up. But they ONLY apply to his set up.
Bensen's spreadsheet for reactive loads gave me the idea for integrating a protection locus onto the SOAR locus to see what the output stage is capable of and where the protection cuts in. However I discovered that the tripping voltage of the protection transistors is critical to the actual limiting current. I have asked for a method to check the actual as built limiting current. This would allow trimming of R28 & R30 to ensure the designed limits are achieved in practice. The last thing I want is premature limiting interfereing with the music and similarly non limiting putting the amp at risk.
the protection values used by Leach and repeated on this thread use a constant current curve when output voltage is less than zero (negative output). This leads to severe overloading of the output stage before the protection cuts in when driving reactive loads.
The ideal curve would be a falling limiting current as the negative output voltage increases.
Using R68 would achieve this but it's value has to be designed as a matched set for the whole protection locus. It cannot be added to an existing set of Leach resistors.
I posted a set of values for the conditions applying to Still4given's set up. But they ONLY apply to his set up.
Bensen's spreadsheet for reactive loads gave me the idea for integrating a protection locus onto the SOAR locus to see what the output stage is capable of and where the protection cuts in. However I discovered that the tripping voltage of the protection transistors is critical to the actual limiting current. I have asked for a method to check the actual as built limiting current. This would allow trimming of R28 & R30 to ensure the designed limits are achieved in practice. The last thing I want is premature limiting interfereing with the music and similarly non limiting putting the amp at risk.
Hi Byrd,
I just plugged your resistor values into my spreadsheet.
One tiny region around 41A and output voltage of 15V is inside the protection locus for the 25degC SOAR. All the rest of the protection is effectively useless, in fact much of it is outside the 100mS SOAR. e.g. 11.8A @ Vce=99.6V. But the safe one second current from the SOAR is only 3.15A @ 99.6V. If you increase Tc for 3 pairs of MJL4302/4281 to 40degC then the safe current falls to 2.77A about 24% of the protection locus your resistor values are predicting.
In case you think that we could be using the 100mS SOAR that is what C17 & C18 do. They increase the short term limiting values to some way beyond the DC values. With very short peak impulse currents the time constant of these caps will allow pulses many times the one second values without the limiting cutting in. This impulsive behaviour will depend on your chosen RC time constant for the protection caps.
I just plugged your resistor values into my spreadsheet.
One tiny region around 41A and output voltage of 15V is inside the protection locus for the 25degC SOAR. All the rest of the protection is effectively useless, in fact much of it is outside the 100mS SOAR. e.g. 11.8A @ Vce=99.6V. But the safe one second current from the SOAR is only 3.15A @ 99.6V. If you increase Tc for 3 pairs of MJL4302/4281 to 40degC then the safe current falls to 2.77A about 24% of the protection locus your resistor values are predicting.
In case you think that we could be using the 100mS SOAR that is what C17 & C18 do. They increase the short term limiting values to some way beyond the DC values. With very short peak impulse currents the time constant of these caps will allow pulses many times the one second values without the limiting cutting in. This impulsive behaviour will depend on your chosen RC time constant for the protection caps.
Hi Jens,
can you clarify the C7 value range.
BOM calls up C7=C12=470uF, C9=180pF
Following my last post. Any ideas for the protection cap values. BOM C17=100pF. Could the value usefully go as high as 10uF?
Do you agree the philosophy that the RC time constants for the input, NFB and PSU should be staggered and that the input should be half an octave above the NFB and in turn another half octave above the PSU? ie RC about 100mS for input, 140mS for NFB and 200mS for the PSU.
can you clarify the C7 value range.
BOM calls up C7=C12=470uF, C9=180pF
Following my last post. Any ideas for the protection cap values. BOM C17=100pF. Could the value usefully go as high as 10uF?
Do you agree the philosophy that the RC time constants for the input, NFB and PSU should be staggered and that the input should be half an octave above the NFB and in turn another half octave above the PSU? ie RC about 100mS for input, 140mS for NFB and 200mS for the PSU.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.