Adding current buffer to pre-amp

For the three and more legged things
brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Adding current buffer to pre-amp

Post by brig001 » Thu May 16, 2013 9:21 am

Hi, still in KL, but contemplating a project for when I get back. I had toyed with the idea of a Pass F5 or a JLH, but they seem over the top since I only need 1 Watt, and I need 4 channels. I came up with this as an alternative. There's no power amp as such, just a class A, unity gain current buffer for the pre-amp. The feedback can be removed from the pre-amp and put on the end of the buffer to get low distortion, or left where it is and have no feedback.

I think of it as a JLH with the middle transistor upside-down, and imagine (no sim software or test equipment) that the top transistor works as an emitter follower with an active load formed by the bottom transistor (one of the definitions of the JLH).

Is this likely to work?

Thanks,
Brian.
Attachments
Amp.png

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Mon May 20, 2013 1:04 pm

Hmm, I wasn't sure either.

Note: The links below are to a Java applet.
Note 2: I have no idea what the first note means :D

I have found this useful: http://www.falstad.com/circuit/ and it even gives a link to my project Project

It sims exactly as I expected, but I have found an issue - I'm not sure how to stabilise the quiescent current. I could make it adjustable and keep adjusting it till it stops wandering (JLH style), but it would be nice to make it automatic. The problem with this (and the JLH) is that the current in the output stage varies with signal, so it's not easy to just measure it. The screenshot attached shows the output and current in each transistor on a scope view to show what I mean.

I'll give it some thought unless anyone has any ideas.

Brian.
Attachments
Amp SIM.png

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Tue Jul 23, 2013 10:16 pm

Time for a bit of an update.
I've decided to build this. It's a diamond buffer to go on the output of the preamp. Should be good if the sims are to be believed: -90dB harmonics at 1V and -130dB at 100mV all open loop (don't know how to do THD in spice yet...). Won't be anywhere that good in practice, but should be good enough. If anyone want more of the story, it's here: http://www.diyaudio.com/forums/solid-st ... reamp.html

Brian.
Attachments
Diamond Buffer3.png

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Tue Jul 23, 2013 10:46 pm

Just figured it out: 0.005% THD @ 1V and 0.00029% THD @ 100mV. Not bad for open loop. The buffer is class A up to about 2W which explains the reducung distortion with reducing power.

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Sat Jul 27, 2013 3:59 pm

Well, I've started building, and the next time I want to build a really small, class A amplifier with no heatsink, please will someone give me a slap. It just fits as you can see. I've also found room for a common mode choke for the output. The theory being that the choke works to block RF from coming back up the speaker leads, but no no one puts a choke in the ground. Well I have. The outer winding has the highest inductance (about 2uH) and that is in series with the output. The inner winding has a lower inductance (about 1uH) and that is in series with the ground lead. They should work together for RF and give a higher inductance, and cancel for the audio.

Anyway, got to figure out how to test it then build the other 3 :(
Attachments
IMG_1770.JPG

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Sat Jul 27, 2013 4:09 pm

Just to clarify, yes that is a penny. Yes it is a new (post 1971) one. The board is 72mm x 47mm or 2 7/8" x 1 7/8" in old money.

2W class A, power dissipation = 0.28W per transistor = 5.6W total

JamesD
Old Hand
Posts: 607
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

Post by JamesD » Sat Jul 27, 2013 6:23 pm

Hi Brian,

There is a phenomenon called 'ground bounce' that is caused by inductance in the ground line or ground plane - whereby rf signals generate a voltage across the ground line inductance and cause circuit performance to shift accordingly... this normally is only an issue with ics and fast circuits BUT it can be a problem at audio and with discrete circuits if it happens inside a feedback loop (intentional or otherwise)...

I don't think you have that case here - depends a bit on how the input signal ground is handled but its a 'keep in the back of your mind' thing if you get some unexplained behavior...

I had the problem with a valve amp that I put the some inductors in the ground between power stage and VAS/driver stages thinking to remove the power supply current surges from the earlier stages... all it did was create an instability that came and went depending on the output level and rf noise in the environment...

James

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Sun Jul 28, 2013 11:16 am

Thanks James,

I don't think it will be an issue here since the chokes are in series with the speaker only and after the feedback. I will watch out for it between the pre-amp and this driver stage though as that could cause some fun if I get it wrong.

Regards,
Brian

User avatar
Mike H
Amstrad Tower of Power
Posts: 17103
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

Post by Mike H » Tue Jul 30, 2013 7:59 pm

Whenever I tried RF chokes like that in series with the speaker I lost treble. Just saying is all. Better plan might be Zobel (RC) network between output and ground.
 
"The beer was so flat it could have been served in an envelope...."

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Sun Aug 04, 2013 11:07 pm

Hi Mike, I have a Zobel too - in fact two Zobels too, one before the choke and one after. I think one of my neighbours has a dodgy appliance (ooh er) and I am getting intermittent crackles at the moment, so I'd like to do what I can to mitigate RFI. Even if the chokes don't cancel fully, I'll have 3uH total which has an impedance of 0.38 ohms at 20kHz, so I shouldn't notice too much.

I have a theory that part of the "valve sound" (if there is such a thing) is down to immunity to RFI (lack of feedback etc.), so this should help.

Brian

User avatar
Mike H
Amstrad Tower of Power
Posts: 17103
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

Post by Mike H » Mon Aug 05, 2013 11:40 am

Valve amps are not immune to RFI.

This has no Voltage gain, and yet it's unstable - but then it does have current gain, and possibly too much, afraid I tend to agree with Struth on diyaudio. I've often had instability because of the high base impedance to transistors with no load on the output side - in fact my germanium o/p amp is unstable without a speaker load for that same reason.

Me I'd just use an op-amp, if only becuae all this sort of stuff will have been thrashed out and sorted before it goes to production.
 
"The beer was so flat it could have been served in an envelope...."

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Mon Aug 05, 2013 12:18 pm

Valves aren't immune, but the topology typically used is more immune that that typically used in transistor amps. The global feedback loop on transistor amplifiers can allow RFI from the speaker leads to get back to the input stage.

No, this isn't unstable. It isn't built yet! My existing system suffers from crackles. I'm just trying to pre-empt this and design something in, rather than trying to add it afterwards. This circuit shouldn't be unstable at all as the same circuit is used in an LME49600 buffer for example. I understand where struth is coming from, but I don't want a class AB - it's really hard to get one right. My buffer should stay in class A for the first two watts, and as we have seen at Owston, I only need one watt to fill that room.

With regards to using an opamp, I would if one existed that had reasonable performance, a couple of watts output, unity gain stable and works off a split supply (no output capacitor).

User avatar
Mike H
Amstrad Tower of Power
Posts: 17103
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

Post by Mike H » Mon Aug 05, 2013 2:55 pm

My mistake, the way it reads it looks like this circuit has got the RFI problems!
 
"The beer was so flat it could have been served in an envelope...."

brig001
Old Hand
Posts: 390
Joined: Sat Nov 13, 2010 9:56 pm
Location: Back home in Preston now

Post by brig001 » Mon Aug 05, 2013 4:04 pm

I think it's mine. Having read it again, I can see where you were coming from :oops:

User avatar
Mike H
Amstrad Tower of Power
Posts: 17103
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

Post by Mike H » Tue Aug 06, 2013 6:33 pm

OK back to the plot - :D - I'm worried you may get distortion caused by the non-linear nature of the base-emitter impedances of the o/p transistors where the signal Voltage passes through the crossover point (= DC zero).

This is where e.g. a pure class A would be superior.

Conversely a NFB-looped op-amp would 'force' its p-p o/p stage to 'behave'.

Just IMHO.
 
"The beer was so flat it could have been served in an envelope...."

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest