Grid bias - sound quality of battery or PSU?

[andyjevans]

Can anyone quote some actual experiences of changing between battery bias on the grid and the same voltage derived from a power supply?

Also, variations in obtaining the bias from the power supply - chokes, voltage regs, different values and types of capacitor etc.

Looking for a bias in the 20-75v range for driver tubes. Presently using two small 12v batteries for 24v, but want to look at options, since if a battery drops voltage the current through the tube increases.

But basically, sound quality is the issue here in making decisions!

Andy

 

[andyjevans]

Here's an example of negative bias supply with EZ81 and choke input!

http://www.users.globalnet.co.uk/~valveamp/images/SE108IT2.GIF

Nice. The designer presumably listened to it and heard a difference.

andy

 

[Doz]

I did some experiments with NICADs in the cathode of small signal stages. Ended up happier with LEDS. No help really to you though.

 

[andyjevans]

I've been doing some listening experiments:
- 2,200uF cap input supply: treble sounds a little dirty
- 2,200uF+8.6K+220uF+8.6K+220uF: pretty much similar
- 2,200uF+1K+220uF+1K+220uF+1K+47uF polypropylene: audibly better in the treble, but treble still a little dirty though less than before.

I read that the last cap is in the signal path, so should be polypropylene, and this seems to have some truth in it.

Haven't tried chokes yet because these would need a bleeder resistor and it would drop the bias voltage, which is just OK as it is.

andy

 

[sjs]

Here's an example of negative bias supply with EZ81 and choke input!

http://www.users.globalnet.co.uk/~valveamp/images/SE108IT2.GIF

Nice. The designer presumably listened to it and heard a difference.

andy

Yes, I did listen to it, and heard a huge difference, the choke input filter did the same here as in a regular positive HT application. So if you have a preference for choke input supplies, you will prefer this for similar reasons.

The EZ81 means the -ve rail takes time to come on, so care is required over the timing of the HT. GZ37s were even slower, hence their application here. If you change rec tubes consider switch on times, or safety systems etc

Prior to this I had used sand rectifiers with cap inputs, tube recs with cap inputs, then choke input. Doing it today (the schema is v old!) I would probably go LCLC with smaller caps, and drop the 100uF bypass cap in the grid circuit.

I cannot offer direct comparison with battery bias, as I have only recently tried this out on my phono & line stages after reading the 26 preamp thread. Nice!

best of luck
Simon

 

[harrylime6]

Hi, Andy
I would suggest that you purchase a sufficient number of 9v alkaline batteries to meet your bias requirements and then place them in the grid circuit to provide the bias value you seek. There are many who will be appalled by what I am suggesting - but, hey, that's life. If you require more precision than 9v increments will allow, then, adding some alkaline double AA's to the mix and wiring them either with or against the current flow should get you to within a foot of your target voltage. Then it's listening time. Hope this helps.