Let us know when you get started on your build..And as you mentioned we should be hearing some news/results soon from Eric.🙂
Hi guy's,
The amp is running again, I am now running it with a supply rail of 70 volts. That is with the LM's pulling their 1,85 amps. Now one question, how can I measure if a MOSFET is working or not? As I am having some doubts about one of the 4's I have. I will be waiting for Shanex and his signal gen. + expertise to give this puppy a test run to the limit! 😀
Eric
The amp is running again, I am now running it with a supply rail of 70 volts. That is with the LM's pulling their 1,85 amps. Now one question, how can I measure if a MOSFET is working or not? As I am having some doubts about one of the 4's I have. I will be waiting for Shanex and his signal gen. + expertise to give this puppy a test run to the limit! 😀
Eric
Hi Andrew, agree with you there. The original circuit didn't have any resistors there but its easy enough to add. Try something around 0.5ohm or 1 ohm.
Thank you guy's for the tip on the resistors🙂 It was applied and both transistors work just fine. When I will put this puppy in it's final casing, I will do one mod, and it will be to replace the 2 separate resistors at the gate of each MOSFETS by 1 equivalent resistor for both. (Ref: second schematic posted earlier on this thread)
Eric
Eric
Last edited:
do you mean r13 & r14 in fig2 of post1?
No!
these gate stoppers must be fitted very close to the gate of each mosFET.
Each gate needs it's own very low inductance connection to it's own gate stopper.
No!
these gate stoppers must be fitted very close to the gate of each mosFET.
Each gate needs it's own very low inductance connection to it's own gate stopper.
If it is the case, can you please explain to me why on the schematic of post 22 (same MOSFET, with same op amp similar power output), there is only 1 resistance R-4 for both MOSFETS?
Eric
Thanks for the advice, yes there are some things in the article that seem ambiguous.
Eric
I post it here:http://www.diyaudio.com/forums/solid-state/179475-its-only-need-50v-45v-outputs-3.html#post2439521
Hi Eric,
At least it is working.
Hi Ontoaba, Thanks for the link to you post. It is a very interesting design (further discussions to your Post..when I get a chance..)
Hi Eric, Looks like the mods on the soft start worked🙂 Now that the HVKs are happy...next step
IMHO: Magic does not exist. It is an effect that is sometimes based on an aberration of perceptions and ,perhaps, a misassumption of what is in fact really happening. Another April Hint(!)
All kidding aside, the published design of the amp did not go into a lot of technical details, and having read the article (in my sometimes limited French) there were (many) issues which needed to be addressed to do a proper build. But I think the article was written towards more of a ''pastime'' hobbyist audience.. there is a bit more going on with the circuit that would be beyond the scope of explanation to their magazine readership. That is definitely NOT the case with diyAudio members!!(I say that with complete respect) as can be seen by the numerous suggestions and comments.
I had a chance to listen to the amp, and (even) at the reduced bias/power level the amp sounds very good (IMHO) and the active current sinks (LMs) do not seem to alter or colour the sound.
I think worth going to the next step of increasing the power level...time for a bigger transformer 😱
That's all from me..I'll leave it to Eric to update everyone on the status and results
A little update,
After "making my own" 1Khz signal, thank you to the following programs.
1) "The Audio Test File Generator"
(Thank you Mr. Taylor)
2) Audacity: Free Audio Editor and Recorder
Audacity was used to cut and paste the 120 seconds signal until I had a 30 minutes version.
Then I burned a CD and Voila! my own home made signal generator!
So, biased at 1,82 Amps the amplifier gave the predicted +- 13 watts. Of course the next step is to increase the bias until we get to the famous 70W. But before I do I have to change my transformer because the one I have (been a recycled one) it is pretty much at the end of it's useful life! Also I will set the LM's as directed in the article because we think it is not only about smoke and mirrors. ShaneX believe there is a little devious trick used in this design to increase the bias in a "automatic" way. And since I have gone so far as making this puppy work, I might as well take the next step upward😀
Anyway I will keep you updated latter, as It will take a couple of week's until I get my hands on my new transformers.
Eric
After "making my own" 1Khz signal, thank you to the following programs.
1) "The Audio Test File Generator"
(Thank you Mr. Taylor)
2) Audacity: Free Audio Editor and Recorder
Audacity was used to cut and paste the 120 seconds signal until I had a 30 minutes version.
Then I burned a CD and Voila! my own home made signal generator!
So, biased at 1,82 Amps the amplifier gave the predicted +- 13 watts. Of course the next step is to increase the bias until we get to the famous 70W. But before I do I have to change my transformer because the one I have (been a recycled one) it is pretty much at the end of it's useful life! Also I will set the LM's as directed in the article because we think it is not only about smoke and mirrors. ShaneX believe there is a little devious trick used in this design to increase the bias in a "automatic" way. And since I have gone so far as making this puppy work, I might as well take the next step upward😀
Anyway I will keep you updated latter, as It will take a couple of week's until I get my hands on my new transformers.
Eric
Hi everyone,
I have been meaning to post an update on this built for the longest time, problem, tooo busy!
Anyway, I tough some of you might be interested in knowing that there was way more to this amp that what we first tough ShaneX and I. But first I will refer you to this Datasheet:
http://www.ti.com/lit/ds/symlink/lm117hv.pdf
On page 6 there is a current limit graphic. As we found out, (after a lot of trials) the amplifier will give its 70W output with 1.82Amps (or so) of idle current. How is that possible? Magic? Nope, simply follow the set-up indicated in the article ie: run the LM117's with 0.82ohm resistors.
At first I tough, why go to the minimum rated resistor to get the required current? Might as well put more regulators to split the load, It would put less stress on each regulator. It turns out I was sooo wrong as even with more current it was still set-up as a common collector amplifier, so it did not give the intended result. After checking other design of the same writer, we notice he was always using the same 0.82ohm resistors to set the current. I tough it was "weird", ShaneX found the answer...The reason for this resistor is to set the LM117 in current limiting mode, if you prefer, it will increase and decrease it's current based on the voltage between it's input and output. So higher voltage less current, less voltage more current. It is actively following the curve of the previously mentioned graph.
Quite frankly I am impressed, this is clearly thinking outside of the box! And I tough it would be worthwhile to share with you guy's who followed this thread.
Cheers,
Eric
I have been meaning to post an update on this built for the longest time, problem, tooo busy!
Anyway, I tough some of you might be interested in knowing that there was way more to this amp that what we first tough ShaneX and I. But first I will refer you to this Datasheet:
http://www.ti.com/lit/ds/symlink/lm117hv.pdf
On page 6 there is a current limit graphic. As we found out, (after a lot of trials) the amplifier will give its 70W output with 1.82Amps (or so) of idle current. How is that possible? Magic? Nope, simply follow the set-up indicated in the article ie: run the LM117's with 0.82ohm resistors.
At first I tough, why go to the minimum rated resistor to get the required current? Might as well put more regulators to split the load, It would put less stress on each regulator. It turns out I was sooo wrong as even with more current it was still set-up as a common collector amplifier, so it did not give the intended result. After checking other design of the same writer, we notice he was always using the same 0.82ohm resistors to set the current. I tough it was "weird", ShaneX found the answer...The reason for this resistor is to set the LM117 in current limiting mode, if you prefer, it will increase and decrease it's current based on the voltage between it's input and output. So higher voltage less current, less voltage more current. It is actively following the curve of the previously mentioned graph.
Quite frankly I am impressed, this is clearly thinking outside of the box! And I tough it would be worthwhile to share with you guy's who followed this thread.
Cheers,
Eric
I'll make two assumptions. This amp is push-pull and we have an 8 ohm load.
With 1.82 Amps idle current you will get only 53 Watts RMS.
Anything other than that is not class A.
With 1.82 Amps idle current you will get only 53 Watts RMS.
Anything other than that is not class A.
Last edited:
It is still a class A as the Mosfet (IRF-150) conducts 360 degree. It is a kind Push-pull as the bottom component will apply a variable bias. However, since the bottom component is not another Mosfet (It is a Voltage regulator) it will follow it's own internal "biasing schedule". So the standard Push Pull formula is non applicable here.
It took us a while to figure it out since there where no clear explanation in the text.
Cheers,
Eric
Oups! 
Reviewing my notes I found that 1.8 was set to test wether or not the MOSFET would survive. The final current was around 2.4, I say around because there was some variations between regulators when pushed to the limit!
Cheers
Eric
Reviewing my notes I found that 1.8 was set to test wether or not the MOSFET would survive. The final current was around 2.4, I say around because there was some variations between regulators when pushed to the limit!Cheers
Eric
- Status
- Not open for further replies.