811 A2 SET
Hi guys,
I've coming out of about a decade of hibernation, been thinking very hard about building a 811a amp after seeing some friends build theirs, and understand the entire challenge with grid current draw.
I have some concerns with regards to my design and I hope to hear some opinions.
1) 811a operating points
I understand that a lot of designers have picked this point, and apparently it sounds good. 430 - 450v @ 85ma - 100ma Iq. Is there another point worth investigating?
2) Is PSRR on the source follower going to be an issue? and if so, is a Jung super reg on the 60V supply going to help? Also, is 60V enough headroom?
3) Is the simulation of 19ma Ig at 450v/85ma a fair assumption?
4) would replacing R5 with a crude CCS improve the source follower significantly.
Thanks!
Hi guys,
I've coming out of about a decade of hibernation, been thinking very hard about building a 811a amp after seeing some friends build theirs, and understand the entire challenge with grid current draw.
I have some concerns with regards to my design and I hope to hear some opinions.
1) 811a operating points
I understand that a lot of designers have picked this point, and apparently it sounds good. 430 - 450v @ 85ma - 100ma Iq. Is there another point worth investigating?
2) Is PSRR on the source follower going to be an issue? and if so, is a Jung super reg on the 60V supply going to help? Also, is 60V enough headroom?
3) Is the simulation of 19ma Ig at 450v/85ma a fair assumption?
4) would replacing R5 with a crude CCS improve the source follower significantly.
Thanks!
Last edited:
1.) Without knowing OPT parameters (at least primary DCR and reflected impedance) the B+ and suggested current is only assumption.
With these -suggested- parameters the tube near -or more- to the dissipation limit.
Be sure, that this not more than 80%, even the tube will be short-lived.
Loadline calculator for power stages with reactive load - Vacuum Tube Amplifiers - DIY
2.) 60V is insufficient, the "border" is at +60V, so 20-30V headroom is requireing.
3.) 450V not the tube anode voltage!
Rather use 430V, 85mA, -22.5V operating point.
If you see datasheet, approximately 15..20mA grid current expected.
4.) Without -30..40V negative supply it's impossible.
With these -suggested- parameters the tube near -or more- to the dissipation limit.
Be sure, that this not more than 80%, even the tube will be short-lived.
Loadline calculator for power stages with reactive load - Vacuum Tube Amplifiers - DIY
2.) 60V is insufficient, the "border" is at +60V, so 20-30V headroom is requireing.
3.) 450V not the tube anode voltage!
Rather use 430V, 85mA, -22.5V operating point.
If you see datasheet, approximately 15..20mA grid current expected.
4.) Without -30..40V negative supply it's impossible.
Last edited:
1) I forgot to mention, I am looking at either a 5k or a 7k OPT, I am leaning towards a 5k one, though i have heard great results on a 7k PriZ.
I did a quick plot (again) and noticed that 100ma op point is very near the 45w limit, looks like it may be valuable to come down to 430v @ 85ma.
I assume to looks 10-15v in the OPT, I haven't got them on hand yet, but i think it is a fair assumption?
2) so 60AC to a modified Jung super reg that will put out 60VDC (V3) that would work? or you think V3 should be even higher to buffer for swing. After some simulation, i realized the 811a has pretty high Mu, it appears only need to swing the grid about 40v pk-pk to get full power.
3) great advice, thanks
4) would a simple TL783 suffice? still need an negative supply or just more headroom? is this assumption based on that the tube is biased to +22.5v and if the signal swings the grid -20v pk, i will be only left with 2.5v of headroom for the CCS to work?
Thanks a million
I did a quick plot (again) and noticed that 100ma op point is very near the 45w limit, looks like it may be valuable to come down to 430v @ 85ma.
I assume to looks 10-15v in the OPT, I haven't got them on hand yet, but i think it is a fair assumption?
2) so 60AC to a modified Jung super reg that will put out 60VDC (V3) that would work? or you think V3 should be even higher to buffer for swing. After some simulation, i realized the 811a has pretty high Mu, it appears only need to swing the grid about 40v pk-pk to get full power.
3) great advice, thanks
4) would a simple TL783 suffice? still need an negative supply or just more headroom? is this assumption based on that the tube is biased to +22.5v and if the signal swings the grid -20v pk, i will be only left with 2.5v of headroom for the CCS to work?
Thanks a million
1.) Not many 7k SE transformer tolerate 85..100mA (too much turns, too much excitation).
It's not essential difference of power between 5k and 7k load, because B+ is relatively low, so I suggest to use 5k.
2.) At least 80-90V DC voltage requiring for SF. Why not divide it from B+?
4.) IMHO there CCS is unnecessary. If use it, must to provide enough headroom for device, so must to use negative voltage too.
It's not essential difference of power between 5k and 7k load, because B+ is relatively low, so I suggest to use 5k.
2.) At least 80-90V DC voltage requiring for SF. Why not divide it from B+?
4.) IMHO there CCS is unnecessary. If use it, must to provide enough headroom for device, so must to use negative voltage too.
1) cool, fair point, was worried that the Rp of the 811 is very high, and would benefit from more Z
2) I am concerned about the amount of heat this will create, in addition to the "high" current requirement of 30+ ma per channel. i figured it might be more efficient to get a separate supply.
4) fair point.
2) I am concerned about the amount of heat this will create, in addition to the "high" current requirement of 30+ ma per channel. i figured it might be more efficient to get a separate supply.
4) fair point.