High ZSAC on Tetrode amp

[Naysayer]

Fine tuning homebrew grid-driven 2x 4cx300 amp.
Plate current meter is accurate now .

Specified ZSAC for plate voltage 1500 and up is 100mA. With 2 tubes I should see 200mA.
Bias voltage adjustment can reduce zsac by raising bias voltage for a zsac of 230mA which is still too high.
My memory of Triodes is that correct zsac coincides with correct Bias voltage.

Does tube quality factor into this? Currently, I'm running my worst set of tubes. OK, I'll swap in a set of different 300y's and re-post later today.

 

[Naysayer]

A tube swap reduced the difference between published bias voltage & zsac. Datasheet states -70v grid bias but to obtain a 200mA zsac I need to increase bias to -74v.
So what should I follow? the published -70 with higher zsac or raise bias voltage to -74 for the lower zsac?
These are all old 300Y tubes. I'd like to understand this better before I put the NOS 300A's back in.

 

[Naysayer]

Control grid voltage at -74 (as opposed to -70) working fine, I see no difference in PO. Plate current however, is noticeably lower.

P.S. I'm going to close it up and stay with the -74 bias for awhile see how it goes. Maybe put the newer 300A's back in because I know those tubes better.

 

[nomadradio]

We routinely see significant variation in the exact grid bias needed for a desired ZSAC. That's why a commercial amp has a pot for this. If your bias adjustment won't reduce anode current enough, you just need a higher negative bias max voltage. The bias voltage needed will depend more heavily on screen voltage than on plate voltage.

Our routine test for a 4CX250B is to note the grid bias voltage needed to establish 50 mA, 100 and 200 mA anode current. The variation can be pretty wide. And goosing the screen voltage above the textbook 250 Volts will make a higher negative control-grid bias necessary to compensate.

73

 

[Naysayer]

Thank you for suggesting a zener swap, I will do that in near future. The amp's personality improved a whole lot after the bias adjust. I'm using an old set of 300Y tubes now and they look burned & tarnished but they're giving me the best numbers I've seen yet.
neil, NYC

 

[Naysayer]

I removed old pics because of the part swaps above and below chassis. This is top pic showing new Roller, Tune & Load air variable. Progressively shorting rotary panel switch adds 120pF caps as needed for lower bands. Tune cap needs almost nothing.

I'm seeing 300 mA zsac (2 NOS 4cx300a) while datasheet says I should see 200mA. The PO is very good though. Tube markings in red ink: 8449 and 8421. Stamped into metal: 4NLA1026 and 4EEA0105. Does the 84 mean 1984? (tubes were spotless/not a mark on them) could that affect zsac, breaking in period?

If not, 100mA still sounds too high to ignore. Altering Bias from the datasheet listed -55vdc --either up or down-- decreases performance very fast. Have not yet tried altering Screen voltage but I can go up or down with little fuss. Next power on I will try with different Screen volts.

NY City

 

[UDX-4570]

Wow, you have a beautiful tube linear, I had to comment, I saw the photo of it. If you don't mind me asking, where did you aquire the HV transformers for it???... I've been wanting to build a linear around two 813's but a little stumped on what and where do I locate some decent transformers for it.
Thanks for reading!

 

[nomadradio]

Didn't say what your target anode voltage is, but I'll just pluck 2250 Volts from the RCA data sheets at Frank's tube database: https://frank.pocnet.net/sheets/049/8/813.pdf

To use a full-wave bridge rectifier circuit, a transformer secondary of 1575 would be called for. We'll just round that up to 1600 Volts AC. Simpler.

Current is the next parameter to pin down before ordering transformer. Tube wants a max of a quarter of an Amp, 250mA. The transformer secondary needs to deliver 1.414 times that, since the filtered DC output voltage is multiplied by that factor.

This now calls for a (roughly) 300 mA current rating from the HV transformer. Yeah, I rounded it down 6 percent.

So, if you use a full-wave bridge circuit, this requires a 1600-Volt 300 mA transformer output, more or less.

That voltage times the current shows us you'll need a transformer rated for no less that 500 Watts. More is better only because it will run cooler as the rating exceeds actual current demand.

I have a suggestion based on first-hand experience. The toroidal shape power transformers made by Antek in New Jersey are rugged, and perform as claimed every time. Fitting them into a cabinet might seem a bit clumsy. Until you see the price. No better deals out there, Watt for Watt. Sixteen pounds, a hundred and five bucks plus shipping.

https://www.antekinc.com/as-8t800-800va-transformer/

This transformer has two 800-Volt secondaries. Rather than wire them in series for 1600 Volts AC you could wire them in parallel and use a full-wave voltage-doubler circuit like every Heathkit, Drake, and similar desktop amplifiers.

This only covers the anode-circuit power. You'll still need 10 Volts AC at 5 Amps with a center tap for the tube's cathode, and bias voltages for both screen and control grid.

You're on your own to cobble those together. Time to give my brain a rest.

73

 

[nomadradio]

Hey Neil! Got distracted by UDX-4570.

Just want to say that bias potentials shown on a data sheet are at best approximate.

The tester our shop uses for the more-common 4x150/4CX250-etc tubes has a fixed heater and screen voltage. The control-grid bias is the only variable testing parameter.

We mark the tube's test result with three numbers. The Grid bias to produce 50, 100 and 200 mA of anode current. It's a static DC test, just turns the tube under test into a hot resistor. Just the same this separates weak sisters. Safer for them to go "POP!" in the tester than in a customer's amplifier. The tester was built to tolerate it.

We see a LOT of G1-bias voltage variation, even in new unused tubes. Don't know how much more or less variation you should see with that type tube, but ignore the grid voltage. Choose you ZSAC for the desired anode dissipation with zero signal. I never recommend more than 50 Watts to my customers for the "Idle Power", but that's because too many of them use AM. For sideband you should have more upward leeway on ZSAC. I'll predict you'll hear very little difference varying ZSAC over a two-to-one range. ''

73

 

[Naysayer]

Was too beat yesterday to make it online. Nomad you were right about my Screen voltage causing of high zsac. I must apologize apparently, I forgot I had left the Screen V at 400 when I swapped the 300Y tubes for the 300A's which run at 350v. A senior moment already? I just turned 65. Yet there was method to my lapse: The 300A datasheet states 400v in the "Maximum Values" section of the datasheet while every subsection (differing class of operation AB1, C, etc) states 350v. It was my intention to try 400v on the 300A's apparently, I forgot I left it at 400v. One less adjustment when swapping tubes. Why so much tube swapping? My 300A's are 'fresh out the can new' while the Y's are old so no big loss if ruined when testing something different.

Anyway, when decreased to 350v, the zsac was fine at 200 again. So I'm ready to finish setting the pi Loads cap values for 5 Ham bands. The larger Air Variable Tune cap in pic is broad enough for 10-80 but the Bud Load air variable is only 20-120pF (that is why I needed 120pF ceramic caps) so I padded the Load via the progressively shorting front panel switch (band switch sort of) following the roller (1st roller too small and it allowed me to scrap that whole band switch setup). Ready to finish that process now.