Also, just to clarify, on your PCB, is this in reference to the second 2200uF in parallel too achieve 4400uF as opposed to Blomley's C7 2500uF?
I don't see a 200uF. I thought the feedback network was from Tr11/Tr12 to R4/R5 in the Blomley schematic.
I'll shut up now.
Finding the best parts of the idea.
Very interesting amplifier. But I think it helps to redraw the schematic for understanding (attached). So far, it seems to me that this moves the switching action to the driver (or pre-driver) stage as others have done. In this case, the 10KHz THD is not that impressive. I think the same thing can be achieved with better, more modern topology, and better results. For example, consider a "blameless" circuit where the drivers switch but the OPs do not.
Very interesting amplifier. But I think it helps to redraw the schematic for understanding (attached). So far, it seems to me that this moves the switching action to the driver (or pre-driver) stage as others have done. In this case, the 10KHz THD is not that impressive. I think the same thing can be achieved with better, more modern topology, and better results. For example, consider a "blameless" circuit where the drivers switch but the OPs do not.
Attachments
square wave source
I picked up a trick from Ian that makes a better square wave source, ie no DC offset.
I picked up a trick from Ian that makes a better square wave source, ie no DC offset.
Attachments
Never argue "noise" until you know if the writer means hiss, hum, buzz, scratches, or Martian Music-Radio.
Balancing may work great at reducing waves from Mars.
Never argue "noise" until you know if the writer means hiss, hum, buzz, scratches, or Martian Music-Radio.
Balancing may work great at reducing waves from Mars.
Isn't that exactly what Jan is saying in his post?
"What is does largely cancel is RF and other external noise radiated or induced into the lines, provided that both lines have the same source- and receiver impedance. That is the requirement for balanced line noise reduction: that both lines have the same impedance to ground"
And Jan is absulutally right.
To minimum would mean the sweeping connection (centre pin of the trimpot) to the diode. Which gives you 500 ohms across the trimmer and minimum current.
Quiescent current is through the output transistors, and is measured across either of the 0.5 ohm resistors. The current going through the drivers via these resistors is considered to be negligible.
Of course Jan is right, IF by "noise" DCESpeedy means "random hiss". If DCESpeedy means power supply ripple, balanced may work very well. Not all non-specialists know the difference.
I dunno; if you have a hum problem on your signal, how could balanced help to get rid of it? I don't see it. Inverting the signal to make the other polarity wouldn't work because it also inverts the hum on the signal.
For the electronics, the hum is just another signal component and there's no way except sharp filtering to get rid of it. In such cases, I would suggest to clean up the amp to get rid of the hum in the first place.
Jan
For the electronics, the hum is just another signal component and there's no way except sharp filtering to get rid of it. In such cases, I would suggest to clean up the amp to get rid of the hum in the first place.
Jan
I finally got this working. I need more. I was just doing a quick test and got lost in the music and didn't have a heat sink. First it was Elliot Smith then Lil Wayne then Radiohead and I went away for 45 seconds to get my DAC and it was dead. Can anyone suggest what I probably fried? I will go find a big chunk of aluminum tomorrow I promise. I need to hear it again. It was amazing. Especially what they say about the silence. My apologies. I'm a slightly functioning autistic.
I'm still randomly replacing components trying to get mine to work again. It's not going well 
Maybe it never worked and I was hallucinating.
I thought Blomley's design was a single supply. I'll give up at some point. For now, I would advise people like me who don't know anything to avoid getting sucked into this project. I'm still trying to find an audio forum geared toward DIYers.
"Change the 200uF to 470uF" they said.
(Looks for a 200uF for 20 minutes, doesn't find one)
Maybe it never worked and I was hallucinating.
I thought Blomley's design was a single supply. I'll give up at some point. For now, I would advise people like me who don't know anything to avoid getting sucked into this project. I'm still trying to find an audio forum geared toward DIYers.
"Change the 200uF to 470uF" they said.
(Looks for a 200uF for 20 minutes, doesn't find one)
Hang in there, buddy. It's worth the trouble when you finally hear that creation singing.
Yes, it's a single supply. I see you've asked about capacitor values and voltage ratings a few days ago. For values, use Blomley's original (or the nearest current "preferred" value i.e. 250uF becomes 220uF). For voltage ratings, just look at the circuit and figure out the peak voltage at that point and add 10%. It's not rocket surgery.
Hang in there, buddy. It's worth the trouble when you finally hear that creation singing.
Thank you for the kind words of encouragement!
I'm going to build another one but with a proper heat sink so I don't kill it again.
I'm going to use the ZTX453 and ZTX553 as well as the BD139/BD140 for Q1/Q8.
I've looked through digikey but I'm still not sure exactly whether there's a suitable part there for to replace the NJW0302G/NJW0281G (which I can only find on eBay).
@egra mentioned "bass response" and a capacitor that doesn't exist in either his schematic nor Blomley's. That's what I was asking about.
This is the digikey BOM I have going. Very interested in suggestions.
Blomley amp tsv digikey * GitHub
It's currently US ~$45.
Blomley amp tsv digikey * GitHub
It's currently US ~$45.
Hi Jan,
Are we talking about the same thread? This thread? Whose thread is it and how was it hijacked? What is happening.
(Edit: I think you might have me mixed up with @Astaro / post #111 above. I have nothing to do with that. I was referring to @egra's gerbers / schematic.)
Last edited: