112a preamp need comment please

[milen007]

hi guys

i just build a 112a preamp and would love any comment, idea, critics. please...

i attached the schematic.

however my concern that it drift from the datasheet

my concern is: with my operating Va=152va. According to 112a datasheet, Vg should be around -10v with auto bias. but the actual circuit i got Vg= -15v.

i tried 3 different brand of 112a tubes and they are all around Vg= -14.4 to 15.3v

anything that i did wrong? it does sing well though. or its should be ok and i could leave it like that?

thanks in adv

Erwin

[Vinylsavor]

Hi Erwin,

it is not unusual, especially with old NOS tubes that they can have higher than 100% emission. However in your case it is quite a bit above. Is your filament voltage on target? Too high filament voltage causes excessive emission.

Best regarsd

Thomas

[milen007]

hi Thomas

thanks for the reply. its below the rated indeed. one is 4.5v and the other one is 4.8v. havent fix this yet. should be able to get both to 4.8v.

but with both filament of different voltage. the grid voltage is roughly 15v as well.

do you think i should opt for fix bias?

erwin

[Vinylsavor]

Hi Erwin,

how did you connect the filament supply? I assume it is DC.
If the - side of the filament supply is connected to the filament terminal with the cathode resistor, the bias voltage is increased by 2.5V. If it is connected the other way, the bias is decreased by 2.5V. I assume that is the case, that would explain the higher current, because your bias in that case would actually be 12.5V

Thomas

[milen007]

hi thomas

yep thanks you. you got the points there. i should check on this. yeah the drift is 5v which is the filament voltage. maybe they are some how related

i am using rod's regulator and connect them as instructed. which is positive side of the filament regulator to the cathode resistor pins. and the regulators are floating.

[Vinylsavor]

Hi!

The datasheet values are always given with relation to the mid point of the filament. that means the bias voltage is increased or decreased by half the filament voltage, depending on how it is connected.

The bit of extra current will not hurt the 112A. I would leave it as is

Thomas

[Rod Coleman]

Hi Erwin,

I don't think there is any problem, either!

Normally, using the filament + terminal for the cathode resistor is best - but you can try for yourself and connect the cathode resistor to "filament -". Recheck the cathode voltage - and expect it to be LOWER (ie, lower anode current).

You can run it either way around - let your ears be the judge. BUT, be sure to check that they are connected the same on both sides - or the sound will be very odd.

[Rod Coleman]

Erwin, you asked me also how to use the 112A in the filament bias configuration - like Andy's #26 circuit.

The big difference if you use filament bias - you MUST connect the negative terminal to the cathode resistor (or the bias will not work).

Hopefully, the drawing shows all you need to make this circuit.

As Thomas describes, the effective bias voltage is the mid-point between the filament terminals, so in this case (with filament -ve connected to cathode), we only need to generate about 10V in the cathode circuit. As shown, a 39 ohm wirewound, that can comfortably handle the 2.5W dissipation, is recommended. Unless you have a favourite exotic type, I recommend the Welwyn W22, W23, W24 series of vitreous enamelled wirewounds.

The 250mA Filament regulator will work like this, but capacitor C1 needs a higher voltage rating (47uF, 35V Panasonic AM audio series), and R8 changed to be 1,8K 0,5W.

Finally, note that filament bias can slightly reduce the effective gm of the DHT. Bypassing a 39R resistor would require a large and unpleasant electrolytic, so you will not want to try that, I suspect. For instance, if you operate the 112A such as to achieve 1,8mA/V - a 39R in the cathode reduces it to 1,7mA/V, so you get a very small drop in gain.

Note that with, high gm, or high Rk, this gain loss effect is important to be aware of.

[andyjevans]

Hi - yes, filament bias will work nicely.

The dissipation of the 39 ohm resistor is 2.5W. I would use a higher wattage than 6W, because experience tells me this gets very hot. If it's inside the chassis then the whole of the inside gets hot including the top plate. I'd go for 12W or even more.

I found that using choke input in front of Rod's regulator made an audible improvement. Plenty of chokes are available from Hammond to do this. You can use a small input cap like 220uF, and main cap should be 20,000 to 40,000uF. More is better. The supply needs to be very smooth with filament bias.

andy

[milen007]

hi guys

thanks a lots for the inputs. as usual, all is really appreciated.

its clear now. for the moment i will just leave it as it is. sounds amazing as it better matched my Line output transformer (58H) compare to 26 which has higher Rp

i am using 0.25mH-15mF-0.47R-22mF for the raw.

Rob, i am running the 112a on 4.5Vdc as the transformer cant keep up and the trimpot does nothing to change the voltage at the heater pins. would you think its ok ? or i need to get another transformer?

Erwin