# Parasitic Suppresser

Discussion in 'Amateur Radio Amplifiers' started by KA5ROW, Jun 13, 2012.

Not open for further replies.
1. ### AG6KGuest

 VHF suppressors do not have a bandwidth since they are not a tuned circuit.

The optimum values of R and L depend on the dissipative ability of the VHF suppressor. In other words there are no such values.

Anode R-Load at the frequency of the anode's parasitic resonance is what controls gain at that frequency. The calculation is not simple since the variable Tune-C is in series to gnd.

2. ### G0HZUQRZ Member

Yes but the suppressor is part of a system and by bandwidth I'm referring to over what range of frequencies the suppressor design is most effective for a given system.

At any one spot frequency the calculation is simple because you can model EVERYTHING after the suppressor as a 1 port linear device. So at the problem frequency it will typically be an inductor in series with a small resistance.

Parasitic suppressor should offer minimal losses at the highest operating frequency of the amplifier, while having enough impedance to
to prevent self oscillation at the parasitic frequency. It would seem the trick is to find an optimum compromise that effectively meets
both goals.

Pete

4. ### G0HZUQRZ Member

Hi Pete,
Yes, that's true.

I was referring to a case where I had chosen L and then designed for optimal R for optimal reduction in anode impedance for a given amplifier topology.

But yes, the Pdiss in the suppressor at 10m will have to be considered. My initial thoughts on this are it will be very difficult to 'define' Pdiss at 28MHz because it will be heavily influenced on the capacitance sharing between the first cap in the Pi tank and the internal capacitance of the tube itself. The suppressor sits in between them so up at 10m I'd expect that this ratio of anode capacitance to Cplate cap to heavily influence how hot the suppressor gets.

As the operator can influence this with only minor changes in tuning I would expect that part of the tradeoff requires the operator to keep the plate cap at a sensible value when tuning.

I'd better also make it clear that I'm not trying to suggest it's possible to design 'a one size fits all' RL that is the ideal design for all amplifiers by the way.

Jeremy

Last edited: Jun 25, 2012
5. ### G0HZUQRZ Member

Hi
First of all these are only what I predict will happen as I've never ever touched the insides of a tube amplifier to measure this. I'm just basing this on basic physics.

If you looked at an extreme case (on 29MHz) where the anode capacitance was significant wrt the theoretical plate capacitance then you end up with two capacitors in parallel (with wire+suppressor+dc block etc linking them)

If you forget the suppressor is there and just look at the current flowing between these two similar capacitances then I'd expect to see a couple of amps at RF for a high power tube although it obviously depends on the amplifier itself.

So if you insert a typical suppressor between them with ESR of ballpark 4 ohms at 29MHz then you could get maybe 16W dissipated in the Suppressor.

There will be other factors affecting the Pdiss but that's the most obvious one to me.

As for the tuning example I would expect that the above current will vary depending on how well the amp is tuned up so the suppressor could overheat during tuning as I'd expect the current to fluctuate quite a bit even with only changes of a few pF below the ideal value in the plate tuning cap.

i.e. Pdiss of the suppressor could easily double even with only a slightly less than ideal tune for max RF power. So the design of the suppressor should ideally take this extra margin requirement into account?

I was really referring to the act of tuning itself as the operator might not tune the amplifier up quickly and/or perfectly. I don't think it would have to be mistuned very much the 'wrong' way to cause a significant increase in suppressor heating.

Last edited: Jun 25, 2012
6. ### G0HZUQRZ Member

To put some numbers on this I'd expect the tuned up plate cap to be in the ballpark of 10-15pF although I've never measured the capacitance myself. This means the tube capacitance becomes significant so there will be a fair RF current between these two capacitances when tuned up at full power at 29MHz.

Maybe someone can confirm a typical (measured) value for the tuning cap when set at the 29MHz tuning setting?

7. ### G0HZUQRZ Member

Yes, that wouldn't be wise...

At work we use a thermal imager for looking at solid state stuff . It has a decent colour display and can resolve down to display tiny SMD parts but I suppose from a safety point of view you could only use the imager to peek inside a tube amp via a transparent safety window/insulator to prevent any risk of electric shock.

I'm currently using it at home for some research work and it is a very neat little gadget.

8. ### AG6KGuest

 The most dissipation R-supp is at 29.69 MHz using A0, on the high-V tap, with the tank tuned for max P-out.
Tuning the tank otherwise is not only not wise, it reduces dissipation in R-supp.

9. ### AG6KGuest

 The Matsush-ita MOF resistors we use in our parasitic suppressor can run dull red in a dark room and not change value substantially.

10. ### AG6KGuest

 The VHF suppressor that has the least loss at 29MHz is the least effective one above 30MHz Pete.