Traynor YBA-1A no bias voltage @ Standby (no bias tap on PT)

[nevetslab]

I’ve just pulled out my Traynor YBA-1A’s original power supply caps, all the rectifiers, all the soldered-to-chassis terminal strips to get fresh real estate to build a new power supply assembly. I’ve got a clean & compact layout for the supplies, and now have all that on a 4” x 6” plate that mounts to the power transformer with # 8 standoffs. All the supply components mount on individual isolated terminals, each threaded #4-40 studs. All point to point wiring.

A problem with the stock wiring scheme is that there is no separate bias tap.
http://i872.photobucket.com/albums/a...pse96c723e.jpg Member Chris61 provided this jpg image of the schematic. Mine was identical to this, apart from no adjustable bias pot.

And, as this is NOT a capacitor coupled bias supply, there is no bias voltage on the power tubes until you switch out of standby. Using higher supply capacitance on the bias supply (47uF/160V), it charges up slowly from 0V, which means the power tubes would conduct like crazy until the bias settles.

The only simple solution I can see is to move the circuit location of the Standby Switch, so it sits between the first filter stack (now 2 x 100uF/500V cans in series with their 100k ballast resistors), and the rest of the power supply distribution circuit. I’ve added a HT fuse following that first filter and the OT C/T.

I didn’t even go as far as plugging in the EL-34 power tubes, that I’m using at present, seeing that this bias voltage ramping to it’s set point (circuit now having a bias pot added) would pose a problem. I’m still intending to build a separate series screen supply and change from the EL-34’s to KT-88’s on the output stage, but starting with a cleaned up & restored YBA-1A circuit that’s working properly.

Other refinements are using UF5408 3A/1kV Ultra-Fast Soft Recovery rectifiers (2 per leg), BY399P 3A/1kV Fast Soft Recovery rectifier for the bias supply. All the rectifiers are in the far corner of the chassis, completely away from the preamp circuit. Power & ground distribution all come directly from the power supply filters on the new power plate, so no shared power on the grounds (other than V1 & V2 circuits are common). And, that wiring is bundled together so there’s no loop area in the distribution. Single point ground to the input jack common from the power supply ground, with no charging current flowing in it. Full re-cap thru-out. Moved R20 (4.7k) of the PI stage locally to sit with the rest of the parts for that stage, with it going directly to Gnd, instead of the stupid means Traynor used, having it on the Presence control next to the standby switch. And, put the 100nF cap in series with the Presence control so there's no DC on the pot.

[g1]

B+ must also charge up from 0V when standby switch is closed, is bias charging so much slower from increasing bias filter capacitance? Bias filter is now same value as B+ filter, shouldn't they come up at about the same rate?

[Enzo]

B+ cap is not charging through a 150k resistor like the bias cap.

What is slowly? How long does it take to get up to 80-90% of the bias voltage? Monitor tube current as this happens, what is the maximum tube current you see and how long does it last?


If you are concerned, reconfigure it like the YBA-1

[nevetslab]

It was taking 35-40 seconds to get to 80-90% of the pre-set bias voltage. As I'd rather have bias voltage present in Standby, I just moved the Standby switch, putting it between the 1st filter stage and the HT fuse (ahead of all the circuits off that stage). The HT supply comes up to full voltage in a couple seconds or less...I haven't actually stopped to time it.

It's work in progress at the moment. That move with the Standby switch seems like it is a workable solution. I seem to recall this issue had been discussed before, and the preferred state is to have bias not be switched with Standby. I didn't think to make any timing measurements prior to my tearing down the supplies.

I'll take a look at the charging time when I get back to the amp tomorrow morning. Been at it all day, and time to go home to sleep on it.

I thought briefly about the Ampeg VT-22 bias circuit, though that one seems troublesome to set up, if deviating from their stock circuit.

Now I'm looking for what mistake I've made in rebuilding the amp, as it's not passing signal at the moment. No doubt something silly or stupid.

[g1]

Overlooked the 150K, thanks Enzo.
There was a YBA-1A version that used cap coupled bias like VT22, but same type of standby arrangement so I don't think it would get you anywhere. Agree that changing the standby arrangement is your best bet.
The schematic link doesn't work, I think this is same one:

[nevetslab]

Supply Charging rates & chassis images added

Ah, thanks on that link. I haven't come up to speed with inserts into the text fields yet.

After finding my missing jumper, and restoring it, the Traynor YBA-1A now passes signal and running in its’ revised state, having the Standby Switch between the first power supply filter bank and the HT fuse, from which all the plate supplies are fed. HT fuse post mounted in vacated Gnd Rev Switch hole on the rear panel.

I looked at the bias voltage charging rate at power-up. It reaches 80% of the final voltage in 12 Sec, and 90% in 20 Sec. Now that there’s bias voltage present during Standby, coming out of standby, there’s a momentary plate current overshoot of about 40% for about 200mS, then it drops down to 80% of nominal set plate current (40mA per tube) in about 4 Sec. No turn-on thumps heard in the speaker. A very slight drop in Plate Voltage on the OT when coming out of Standby….about 4 line cycles, dropping about 7%, and recovers fully in 100mS.

The Screen Supply voltage ramps up to 90% of full voltage in 100mS & at full voltage in 200mS, while the PI & preamp supply voltages ramp up to 80% of full voltage in about 4 sec.

[KevinOConnor]

Hi Guys

If you must have a standby, it would be better positioned as a cathode lift or as a screen disconnect. Where it is now, you should use diodes to block the CD and maybe place a small cap (10nF) across the contact. Breaking DC is not what most switches are designed to do to withstand. There is actually not much point in disabling the plate supply or in leaving part of it powered.

The stock standby in these amps lifts the CT of the PT plate winding. This breaks the main charging path for the plate filter caps and also for the bias supply return, so both voltages are lost in standby which is is of no concern. The supplies within most amps track each other well enough under loading, with mains variations, and at turn-on that the tubes are safe and operate fine - at least as intended.

Have fun