Switch at the peak helps with transformer inrush. This can be significant with large toroids. This wont fix having to charge the capacitor, but it does mean the transformer isnt trying to charge it as much right at the instant of turn on. The peak is where dv/dt is lowest, so your doing less transforming at that moment.
andRookie said:
With a delay on B+ no fuses blows. The bias - level pot has 90kOhm to ground to make sure bias cannot be turned off too much.
With a passive pot between a cd player and input there is almost no hum or noise at all. So far I am very happy with the sound.
That is really nice work. Congradulations on the build.
Thanks for all your compliments
As mentioned the schematics is from the well known Byrith 4-30 on the Lundahl pages, but the PSU part is different. I do have a PSU chart I made in powerpoint but I do not know how to upload it.
The Hashimotos gave us (I did the work under heavy guidance from jane and others) another challenge when the amp became unstable with feedback. The coupling capasitors value was lowered from 0,47 uF to 47nF and it helped. The amp now works fine.
EC8010: - the fuse broke when bias voltage - psu, and B+ - psu was turned on by the same switch. In PSUD2 I could see my bias - psu filter took several seconds to reach full voltage. I put a delay on B+, so that bias - voltage on the EL34 grids was ok when B+ switched on, and then the fuse was ok.
As mentioned the schematics is from the well known Byrith 4-30 on the Lundahl pages, but the PSU part is different. I do have a PSU chart I made in powerpoint but I do not know how to upload it.
The Hashimotos gave us (I did the work under heavy guidance from jane and others) another challenge when the amp became unstable with feedback. The coupling capasitors value was lowered from 0,47 uF to 47nF and it helped. The amp now works fine.
EC8010: - the fuse broke when bias voltage - psu, and B+ - psu was turned on by the same switch. In PSUD2 I could see my bias - psu filter took several seconds to reach full voltage. I put a delay on B+, so that bias - voltage on the EL34 grids was ok when B+ switched on, and then the fuse was ok.
Hi andRookie, What are you using for the B+ delay? (Very nice high-end looking amp by the way)
Yugo – That is a useful PDF although it doesn’t say in words that the magnetising current is 90 degrees out-of-phase to the applied voltage (graph A show this though).
EC8010: Core remanent magnetism is perhaps the key. If you are experimenting with point on wave switching ON (watch out for those exploding triacs!) you could investigate if point-on-wave switch OFF could be beneficial – leaving the core remanent correctly polarised for next times switch on.
Has anyone used NTC surge suppressors on valve heaters to reduce cold start inrush?
Yugo – That is a useful PDF although it doesn’t say in words that the magnetising current is 90 degrees out-of-phase to the applied voltage (graph A show this though).
EC8010: Core remanent magnetism is perhaps the key. If you are experimenting with point on wave switching ON (watch out for those exploding triacs!) you could investigate if point-on-wave switch OFF could be beneficial – leaving the core remanent correctly polarised for next times switch on.
Has anyone used NTC surge suppressors on valve heaters to reduce cold start inrush?
Chris - I used Power supply delay action board for the purpose. 6,3V input from the separate heater transformer and one switch for each input on my full wave rectifier diodes for B+. I got 17 seconds delay on B+ so the tube is both somewhat heated and has full bias voltage.
Working on another problem, still its the Byrith 4-30 schematic.
The anode of Ef86 now is at 85V
B2+(ecc83) is 410V
Ra (resisor 150k + trimpot50k) is 175kOhms per triode
Rk is 82k
Vk is 87V (as I remember from the measurment, I`ll check this again)
The problem is that the anodes of the ecc83 is at 280V. They should be at 310V or something. Va=280V indicates that the current through the tube is higher than what we expected. Something is wrong. I`ll measure the current thru the 82K resistor and the actual resistance.
The anode of Ef86 now is at 85V
B2+(ecc83) is 410V
Ra (resisor 150k + trimpot50k) is 175kOhms per triode
Rk is 82k
Vk is 87V (as I remember from the measurment, I`ll check this again)
The problem is that the anodes of the ecc83 is at 280V. They should be at 310V or something. Va=280V indicates that the current through the tube is higher than what we expected. Something is wrong. I`ll measure the current thru the 82K resistor and the actual resistance.
My measurements do not add up!
B2+(ECC83) is 414V. I measure from the filter resistor to ground.
Va(ecc83) is at 285V from ground. All 4 of them give the same result. This should give me a current draw over the anodes of ~0,7mA per triode over the 175kOhm load.
The voltage drop from the filter resistor(B2+) to the anode is from 85 - 90V, some difference between the anodes. This indicates a current of ~0,5mA which is more consistent with the findings at the catode, where Vk=90 and over a 82k resistor for both triodes this gives 0,54mA per triode.
Other measures:
EF86 - Va=86V
Vg(ecc83)=86V
All resistors are also Ohmed and give me the right values.
B2+(ECC83) is 414V. I measure from the filter resistor to ground.
Va(ecc83) is at 285V from ground. All 4 of them give the same result. This should give me a current draw over the anodes of ~0,7mA per triode over the 175kOhm load.
The voltage drop from the filter resistor(B2+) to the anode is from 85 - 90V, some difference between the anodes. This indicates a current of ~0,5mA which is more consistent with the findings at the catode, where Vk=90 and over a 82k resistor for both triodes this gives 0,54mA per triode.
Other measures:
EF86 - Va=86V
Vg(ecc83)=86V
All resistors are also Ohmed and give me the right values.
Re: Inrush CURRENT
If the load is inductive, then you're absolutely right, but I wouldn't expect a typical amplifier power supply reflected through a mains transformer to be an inductive load. I've just ordered a new oscilloscope and current probe at work, so I'll need to check that they work. Perhaps I can get some genuine measurements, rather than hypothesising.
gingertube said:
If the load is inductive, then you're absolutely right, but I wouldn't expect a typical amplifier power supply reflected through a mains transformer to be an inductive load. I've just ordered a new oscilloscope and current probe at work, so I'll need to check that they work. Perhaps I can get some genuine measurements, rather than hypothesising.
IIRC at time zero it does look a lot like an inductive load, once the field is fully established it looks more like a reflection of the load on the secondary. I would be very interested in what you find once you get the right tools to make the measurements.
I have had a lot of problems switching transformers, particularly Rcore and toroids at zero crossing with triacs - resulted in a lot of blown fuses and in quite a few instances exploded triacs as well. (600V/20A triacs for instance) The best fix was to switch at the crest of the sinewave, the next best was to chose a triac controller that did not incorporate a zero crossing detector.
I remember measuring first cycle inrush currents with a current probe which in some instances were well over 70A - this with very large EI type transformers on 120V. (Rcore was even worse) I remember being quite surprised, I would have thought resistance in the ac wiring would have reduced this more than it did. The ultimate solution was an NTC thermistor to reduce inrush to more manageable levels and the aforementioned triac controller.
I have had a lot of problems switching transformers, particularly Rcore and toroids at zero crossing with triacs - resulted in a lot of blown fuses and in quite a few instances exploded triacs as well. (600V/20A triacs for instance) The best fix was to switch at the crest of the sinewave, the next best was to chose a triac controller that did not incorporate a zero crossing detector.
I remember measuring first cycle inrush currents with a current probe which in some instances were well over 70A - this with very large EI type transformers on 120V. (Rcore was even worse) I remember being quite surprised, I would have thought resistance in the ac wiring would have reduced this more than it did. The ultimate solution was an NTC thermistor to reduce inrush to more manageable levels and the aforementioned triac controller.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.