Electronics > Projects, Designs, and Technical Stuff
Voltage sag problems in tube amp (SS rectifier)
TimFox:
If you do an actual analysis, you will find a huge difference between having 100uF or 4uF ahead of the choke. In the heyday of tube amps, one either used a capacitor-input filter or a choke-input filter. The choke-input filter was easier on the rectifier (peak current), but gave less voltage output, with reasonable efficiency and less droop under load.
trobbins:
You have the output stage screens at B+ (ie. circa 500V). I'd suggest the B+ was purposefully designed to sag so as not to damage the screens - that was common for guitar and PA amps. If you want to stiffen the B+ then you need to manage the screens.
Check your schematic against the amp around the input/PI stage cathodes.
You will need to measure the power transformer primary winding resistance, and use the calculator in PSUD2 to get the equivalent secondary winding resistance. Prepare your PSUD2 simulation, including the choke resistance, and effective idle current from the output stage as load, and check that the simulated voltage is close to the measured voltage - and preferably post here as you may make some newb mistakes. Then you can explore what the power supply provides when loading is increased, and confirm that is close to your measured sagged level. Then you can work through any options you may have - and given you probably aren't aiming to replace the power transformer or choke, then that leaves the filter capacitance values, and the repercussions of changing them.
ELS122:
--- Quote from: T3sl4co1l on June 03, 2020, 11:59:06 pm ---Doesn't mean it was designed properly, or at all. Audio is full of "it makes sound, ship it!"
Clearly either the supply voltage or the output transformer is grossly mismatched to the tubes in use. It would be marginal to OK for 6L6GCs, but not 6L6GB/6P3Ss. Or at 16 ohm load if the transformer losses and frequency response are adequate.
Tim
--- End quote ---
while yes many tube amps old and new had useless or just weird parts of the circuit.
the output transformer AGAIN, is 24:1 from one plate to the center tap, not from plate to plate. that would mean that if measuring from plate to plate it would be 48:1
--- Quote from: greenpossum on June 04, 2020, 12:19:30 am ---You say the transformer ran hot when you had a tube rectifier. Maybe this was due to the extra heater draw and the transformer is deliberately underrated because they didn't expect it to be used continuously at full volume hence the droop. Maybe that's just the way it is.
--- End quote ---
AGAIN, the amp was originally CATHODE BIASED. I changed it to FIXED BIAS. which would usually draw MORE POWER.
ALSO, I tried playing the amp trough another outlet in another house and it was nearly ice cold. this would make me think it may just be a bad power factor problem or something.
and ALSO, it is pretty normal for a power transformer to run warm when in use. BUT it's running a bit too warm for comfort. so it's either not made for a fixed bias ~40w amp (which it is not), or it's shorted, which if it was I am pretty sure it would be hot not just lukewarm.
--- Quote from: TimFox on June 04, 2020, 03:04:17 am ---If you do an actual analysis, you will find a huge difference between having 100uF or 4uF ahead of the choke. In the heyday of tube amps, one either used a capacitor-input filter or a choke-input filter. The choke-input filter was easier on the rectifier (peak current), but gave less voltage output, with reasonable efficiency and less droop under load.
--- End quote ---
yeah cause 1. my amp the 4uF capacitor is too small. it resonates at like 30Hz or something, also my choke is 8.4 henries while the one on the marshall amp is 5 Henries. by calculating the resonant frequency, it comes out to 7.1Hz resonant frequency.
--- Quote from: trobbins on June 04, 2020, 03:57:52 am ---You have the output stage screens at B+ (ie. circa 500V). I'd suggest the B+ was purposefully designed to sag so as not to damage the screens - that was common for guitar and PA amps. If you want to stiffen the B+ then you need to manage the screens.
Check your schematic against the amp around the input/PI stage cathodes.
You will need to measure the power transformer primary winding resistance, and use the calculator in PSUD2 to get the equivalent secondary winding resistance. Prepare your PSUD2 simulation, including the choke resistance, and effective idle current from the output stage as load, and check that the simulated voltage is close to the measured voltage - and preferably post here as you may make some newb mistakes. Then you can explore what the power supply provides when loading is increased, and confirm that is close to your measured sagged level. Then you can work through any options you may have - and given you probably aren't aiming to replace the power transformer or choke, then that leaves the filter capacitance values, and the repercussions of changing them.
--- End quote ---
Yes I am thinking about moving the screen supply to a different node,
I already have concluded that I just had bad rectifier diodes.
ELS122:
I put a 100 Ohm resistor in series with the choke (before it but after the 4uF capacitor) and measured 4.7V across is when at idle. so 47mA and the voltage after the resistor was 485V at idle. at full volume it goes up to 15V across the resistor, so 150mA. when at full volume the voltage after the resistor was 385V.
this is with a 5Z4S rectifier.
by using PSUD2 I calculated the source resistance of 130.5 Ohms. turns ratio of 1.6
the load resistance, I have no idea how to calculate it :-\
ok nvm, I figured it out. and even tho it was pretty clear that my diodes were the problem. it looks like it's not really the diodes fault. the voltage should drop about 100V in theory
now how do I fix this?}
nvm x2. PSUD2 is soo useful :D
I found out that if I replace the 4uF cap with 22uF everything is better. the sag is MUCH lower and the ripple too.
also, I drew some load lines, and found out that I forgot to half the output transformer impedance when drawing it before. so yeah with an 8 Ohm load the max output is 8.5 watts class B1 and with an 16 Ohm load it's 27 Watts class B1, on paper. so yeah it all makes sense now. thank's for making me recalculate the output tube output :D and with the power supply problems :-+
TimFox:
One detail about PSUD: the transformer parameters used are the no-load voltage and secondary resistance value, although manufacturers normally specify the voltage at full load. If you know the efficiency, or measure it directly, you can estimate the no-load voltage for the computation. Note that if you use a proper choke without a capacitor before it, the voltage will be lower, but it will not fall so fast with load current. There is a minimum value of current for this effect, for which I refer you to the textbooks (see "critical inductance").
Also, with a vacuum rectifier such as the 5Z4, there is a limit to the capacitance used in a capacitor-input filter, due to the stress limit on the peak cathode current. The 1939 RCA data sheet recommends 8 uF max (unless you have a large source impedance), or 5 H minimum for a choke-input filter.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version