I want to build a small power amp with THF-51S. My plan is to reuse existing hardware from F5 build, so I will be using the chassis and power supply. Ben's approach in this thread has attracted my attention because it has KISS with good sounding. That was confirmed by all builders.
The main problem is the power disspation to 2 x 100W when biasing is 2.5A to 3A. The standard 4U/300 case is limited to max 2 x 70W = 140 W for two channels. This means the bias current should be 1.3A to max 1.6A without any extra Noctua fan.
The result could be following circuit based on Ben's CCS/mu follower. It is simulating good but I am not sure if it's still SE or just PP?
An extra bonus would be saving on the output coupling cap C3...
The main problem is the power disspation to 2 x 100W when biasing is 2.5A to 3A. The standard 4U/300 case is limited to max 2 x 70W = 140 W for two channels. This means the bias current should be 1.3A to max 1.6A without any extra Noctua fan.
The result could be following circuit based on Ben's CCS/mu follower. It is simulating good but I am not sure if it's still SE or just PP?
An extra bonus would be saving on the output coupling cap C3...
huh.......... ?!?
with limited case chunk, think of follower with choke as load
easily reconfigured PSU for single rail, so you can set decent Iq
now, F5 PSU - arranging two rails in series means 45-ish . sorta low for follower with Mu as load
same PSU rearranged secs/rectifier to have one 23-ishVdc rail, still sorta low, but it'll work slightly better than that above
no way Jose to make it without output cap
edit - Botte was faster, same idea, so there must be something
with limited case chunk, think of follower with choke as load
easily reconfigured PSU for single rail, so you can set decent Iq
now, F5 PSU - arranging two rails in series means 45-ish . sorta low for follower with Mu as load
same PSU rearranged secs/rectifier to have one 23-ishVdc rail, still sorta low, but it'll work slightly better than that above
no way Jose to make it without output cap
edit - Botte was faster, same idea, so there must be something
The choke load is a good alternative. I have two choke loaded 2SK180 follower amps and they sound great.
As ZM mentioned, the F5 PS can still be used. With the two rails in series, take V- from V- and V+ becomes Ground.
ZM, I must be missing something but I don't understand why 23V is better than 46V from the F5 power supply if it is used to power the mu follower. Perhaps with choke loading, low voltage and high current may be better than high voltage and low current, or perhaps not.
As for the mu follower follower, it wouldn't be ideal but the lower PS voltage of 46V and Iq will still give you music. The power output would be about half, I would guess, but that would be about the same power output for the choke load with the same power supply. It wouldn't take advantage of the power capability of the THF-51S, but it is still substantial power for a single ended amplifier. It's at least double what you can get from a Sony VFET. The mu follower mosfet can also be a lower power device.
The great thing about diy is that you can try things for yourself and find out for yourself what works and what doesn't work. So my thought is that you can give it a try if you are in the diy mood, for the education and entertainment.
As ZM mentioned, the F5 PS can still be used. With the two rails in series, take V- from V- and V+ becomes Ground.
ZM, I must be missing something but I don't understand why 23V is better than 46V from the F5 power supply if it is used to power the mu follower. Perhaps with choke loading, low voltage and high current may be better than high voltage and low current, or perhaps not.
As for the mu follower follower, it wouldn't be ideal but the lower PS voltage of 46V and Iq will still give you music. The power output would be about half, I would guess, but that would be about the same power output for the choke load with the same power supply. It wouldn't take advantage of the power capability of the THF-51S, but it is still substantial power for a single ended amplifier. It's at least double what you can get from a Sony VFET. The mu follower mosfet can also be a lower power device.
The great thing about diy is that you can try things for yourself and find out for yourself what works and what doesn't work. So my thought is that you can give it a try if you are in the diy mood, for the education and entertainment.
Attachments
I know, I know a choke loaded version... w/ L'Amp and w/o, LD1014, warbler...
Chokes are heavy and expensive, small bias ok, may be a choice but I don't like chokes and transformers unless with valves or tubes.
My question is: What happens when this CCS/mu follower will be powered with symmetrical +/- Vcc = 24V like in my sim?
I still don't know if it is SE or just PP?
Chokes are heavy and expensive, small bias ok, may be a choice but I don't like chokes and transformers unless with valves or tubes.
My question is: What happens when this CCS/mu follower will be powered with symmetrical +/- Vcc = 24V like in my sim?
I still don't know if it is SE or just PP?
If you take the speaker output directly off the SIT source, then the mosfet circuit is a CCS only and the amplifier is all single ended, If you take the speaker output from within the resistance between the mosfet and SIT, then it is a mu follower, and the amount of "assistance" from the mosfet depends on the distribution of resistance between the speaker output and mosfet and SIT.
As for the DC voltage at the speaker output, that can be adjusted to zero with a bipolar power supply as you have done in your simulation. So in theory you can leave out the capacitor. I have not tried a bipolar power supply with no output capacitor so I have no experience with that. I do not know how stable it is, nor do I know how much DC voltage may appear at power-up. In my amp I do adjust the bias so that the SIT dissipates more than the mosfet. For my amplifier, I chose a unipolar power supply for its simplicity, and I have no objection to an output capacitor, and it allowed me to play with the SIT Vds.
Perhaps ZM can elaborate on the output capacitor.
Edit: I see that ZM has commented while I was typing.
As for the DC voltage at the speaker output, that can be adjusted to zero with a bipolar power supply as you have done in your simulation. So in theory you can leave out the capacitor. I have not tried a bipolar power supply with no output capacitor so I have no experience with that. I do not know how stable it is, nor do I know how much DC voltage may appear at power-up. In my amp I do adjust the bias so that the SIT dissipates more than the mosfet. For my amplifier, I chose a unipolar power supply for its simplicity, and I have no objection to an output capacitor, and it allowed me to play with the SIT Vds.
Perhaps ZM can elaborate on the output capacitor.
Edit: I see that ZM has commented while I was typing.
Thx Ben.
A similar approach with a DC coupled F8 called DCSFate with bipolar power supply was successfully resolved by Patrick (EUVL) and other builders (First Watt F8 thread #140). They have confirmed that DC couple is possible even using various output devices IRFP and SJEP and said that there were no surprises and the hardware is behaving quite close to the sim.
If I had a choice between direct coupled output and capacitor coupled, I would always prefer the first version.
I think it's highly recommended to give the sim a try in real but currently my stock of THF-51S is 0️⃣...
A similar approach with a DC coupled F8 called DCSFate with bipolar power supply was successfully resolved by Patrick (EUVL) and other builders (First Watt F8 thread #140). They have confirmed that DC couple is possible even using various output devices IRFP and SJEP and said that there were no surprises and the hardware is behaving quite close to the sim.
If I had a choice between direct coupled output and capacitor coupled, I would always prefer the first version.
I think it's highly recommended to give the sim a try in real but currently my stock of THF-51S is 0️⃣...
Last edited:
difference is THF51 (to said amp)
sole and enough reason why I said you need to confirm it in vivo
there are several things/phases between sim and real shebang, some factors being stability of Iq and DC Offset, same as power on/off behavior
it can be predicted more or less in sim, but as always, proof is in da pudding
sole and enough reason why I said you need to confirm it in vivo
there are several things/phases between sim and real shebang, some factors being stability of Iq and DC Offset, same as power on/off behavior
it can be predicted more or less in sim, but as always, proof is in da pudding
Left channel up and running for over an hour. The chassis doesn’t belong to me - it’s destined for one channel of His Mightiness’s XA252s with 2SK180’s. Once I’d stuffed all the boards, drilled and tapped the M4 holes and dry fitted the XA252 I removed it all and took the opportunity of the owner moving house to steal the use of his chassis. Fitted Ben’s bits and powered it all up in stages. All went well.
Once it cooks for an hour or two I’ll cover it up and recheck temps. Right now, with the mains at 120.3VAC the loaded voltage from my ex-Krell PS is 54.25VDC and temps are 58C on the outside opposite the SIT and 60C opposite the MOSFET.
Next week I’ll wire up the other channel and play music!
And yes, Mightiness, I had good intentions of putting an insulator under all those test cables but….I took a chance!
Thanks, Ben, can’t wait to hear it. And, of course, many thanks to Papa for the endless supply of great amps.
Once it cooks for an hour or two I’ll cover it up and recheck temps. Right now, with the mains at 120.3VAC the loaded voltage from my ex-Krell PS is 54.25VDC and temps are 58C on the outside opposite the SIT and 60C opposite the MOSFET.
Next week I’ll wire up the other channel and play music!
And yes, Mightiness, I had good intentions of putting an insulator under all those test cables but….I took a chance!
Thanks, Ben, can’t wait to hear it. And, of course, many thanks to Papa for the endless supply of great amps.