B+ Ripple - How much is too much??

[Ed/KB1HYS]

According to the handbooks, a class C final can have a decent amount of ripple in the B+, in as much as 5%, and for CW work this would probably be ok. but...

What if that Class C final is being modulated either at the screen or cathode?  Would a fair amount of ripple be heard?

Right now my B+ has about 2.5% ripple. Will this be heard? 

 maybe I could run it off something with 50Hz and "sound exotic"... 

FWIW - modified the power supply to provide 3500V or so. Smoke test to follow sometime this weekend.

Bolted in an old TV power supply chasis & tranny to provide Grid and Screen Bias volts, with a some extra space for relays and such.

Klugged together outta scrap & castoffs. Should work fine! 

[k4kyv]

With plate modulation, the modulation transformer secondary or the modulation reactor adds additional filtering, usually enough to reduce the ripple on the carrier to make it less objectionable on an AM signal.  Ripple is more tolerable on a CW signal  than on a phone one.

[K9ACT]

Right now my B+ has about 2.5% ripple. Will this be heard? 

I am curious to know how you measured that on a 3500 volt supply.

I have heard of a few tricks but none that give me much confidence.

Jack

[WQ9E]

Jack,

For me a Tektronix P-6015A 1000X probe coupled to my 7854 scope.  It is rated for 20 KV DC and the handle keeps you safely away from the action.

[K9ACT]

Jack,

For me a Tektronix P-6015A 1000X probe coupled to my 7854 scope.  It is rated for 20 KV DC and the handle keeps you safely away from the action.

OK but how do you see the 80 mv on top of the 3.5V?  You have just changed the scale but I don't see how you calibrate the two diverse values.

Guess I will have to go try that again but I got nowhere the last time.

js

[WQ9E]

One of the few times that you want to use AC coupling on the scope to block the DC component and just view what is riding on top of it.  

In some situations you can also use a differential comparator plug-in (like the 7A13 for the Tektronix 7000 series) that allows you to offset the DC voltage using its built in source.  This is also a useful plug-in for making accurate voltage measurements in a very high impedance circuit; you use its internal voltage source to balance the source voltage and read the level from the plug-in readout.

[K9ACT]

One of the few times that you want to use AC coupling on the scope to block the DC component and just view what is riding on top of it. 

Seems like I was having a brain cramp here.  I can measure the DC voltage with the probe on a DVM and just use the scope and probe for the ripple.  Seems too simple now.

js

[WU2D]

I run a bridge and a simple C filter (33 uF combined at 1500V ) on my ART13 HV and nobody complains. I am sure I have a lot of 120 Hz ripple.

Mike WU2D

[KM1H]

I run a 10H choke and 6 uF oil for mine and it sounds humless, never looked with a scope. Something to put on my To Do list; I may be in for a shock!

The big amps and modulators are filtered to beyond overkill.

Carl

[w1vtp]

With today's trend toward higher fidelity AM and extended range speakers the old requirements of 2 - 5% ripple is probably no longer acceptable. I think my old Eldico TR-1 used 2 mfd caps in the 1500 volt PS.  Some older unmodified communications grade receivers didn't hear 120 Hz all that well - don't forget those matching speakers either -- the combination just didn't hear the ripple that was completely acceptable in those days.

Today with the advent of very high values of power supply caps we tend to think more of going really high and using stepstart tech.. to limit the inrush current.  I'm planning on using 6 mfd caps with swinging / smoothing chokes in the classic PS filter for my 813 xmtr.

In the old days, I used to use a HP 400H and a blocking cap to check ripple.

Al

[W3RSW]

With today's trend toward higher fidelity AM and extended range speakers the old requirements of 2 - 5% ripple is probably no longer acceptable.

Yeah, can you hear tone control on hifi AM?
Blow your speakers.

Cone across the room.
  Ears a'bleedin' from the boom.
But I read ur tone's loud and clear,
  Turned my rig right off I fear.

[Ed/KB1HYS]

I'm using a bridge rectifier, 2x 866A and 2x3B28's (only had three of each).
The ripple value I got from calculation, based on the old handbook formulas. I have to work up a load to actually test the powersupply under normal operating conditions.

[K9ACT]

I'm using a bridge rectifier, 2x 866A and 2x3B28's (only had three of each).
The ripple value I got from calculation, based on the old handbook formulas. I have to work up a load to actually test the powersupply under normal operating conditions.

I thought that number looked very familiar and too neat.

Yes.   Time to measure.

js

[ke7trp]

I just went through this with my T368.  All of these hum. They have 4 UF of cap.  Small caps everywhere.  In stock form using the Carbon mic and an R390. You just cant hear the hum they put out. But now, people have wider receivers, better speakers and are picky.

GLobe king had 3 or 4 UF. It has hum.  I increased that to a 12 UF oil can and its silent. 

My Friend has three T3s All of them hum. Its not bad, But I can hear it.  I got my hum down to a very low level by recapping the speech amp and changing the Preamp tube to a 12au7.

I still have hum as seen on the scope. Its comming from the DC supply. Its low. I have not measured ripple. But I can see it on the carrier.  99% of the time, Nobody will notice or mention the hum.  Only guys with Band scopes will sometimes "see" it. 

I think that if you are building something, or have the parts and time, I would overkill the PS so there just simply is no hum. On the T3, Thats going to be 15 to 20 UF range. Otherwise, You are going to have some Smugotrons bitch and moan.

Clark

[Opcom]

When capacitor - coupling the hum riding on the HV supply "directly" to the scope:

Most scopes have a 500V input rating. When the blocking capacitor charges, the voltage can reach a positive value close to the B+ momentarily. When the cap discharges (HV turned off), the same, except a negative voltage.

A solution is to connect one end of the cap to the HV supply and the other end to GND through a 100K or 1M resistor. A 1uF cap works well with the 1M resistor to avoid the time constant attenutating the 60Hz and lower frequency noise. It is also interesting to watch the HV slowly swing around a few volts as the line voltage fluctuates.

After switching on the HV, wait a moment for the cap to charge before applying the probe in order to avoid a large positive pulse from entering the scope.

Before switching B+ off, remove the probe to prevent a large negative pulse from entering the scope.

It is also useful to connect a couple zener diodes back to back across the resistor to clip this voltage at +/- some value, just be sure not to exceed their peak current ratings.

This way you can even look at 80mV riding on the 3.5KV



If you connect this to the modulated HV, you can see all kinds of low frequency artifacts created by the interaction of the mod iron, any decoupling caps from it to GND, and the power supply proper. I mention that because if there are unwanted resonances, those things as previously discussed can cause large excursions and put the iron at risk.

[Ed/KB1HYS]

Thanks Patrick!  Thats a good tip.

[WD5JKO]

Maybe I'm paranoid, but if I were to measure HV ripple, and I didn't have a suitable probe, I'd build a 100:1 voltage divider. This would take a 3500v supply down to 35v. Then I'd either AC or DC couple the scope or DVM input. No way to zorch anything with this approach.

With my luck, I'f I hung a cap from the HV B+ to the scope input, even with a resistor to ground, and a bi-directional zener clamp, that cap would short out resulting in a massive flame-out of expensive equipment.

Jim
WD5JKO

[WQ9E]

I am more concerned about the operator than the equipment.  A probe designed for safe HV usage is pretty big and provides plenty of operator protection.  They do show up used and are a good item to keep an eye out for at the surplus houses.  My Tek 6015 has gotten plenty of use since I found it via ebay 10 years ago.

[w1vtp]

Maybe I'm paranoid, but if I were to measure HV ripple, and I didn't have a suitable probe, I'd build a 100:1 voltage divider. This would take a 3500v supply down to 35v. Then I'd either AC or DC couple the scope or DVM input. No way to zorch anything with this approach.

With my luck, I'f I hung a cap from the HV B+ to the scope input, even with a resistor to ground, and a bi-directional zener clamp, that cap would short out resulting in a massive flame-out of expensive equipment.

Jim
WD5JKO

YES!  That's what we did when we were setting up for measurements on the EICO 720 project.  100:1 is a good number.  Be safe and be sure the resistors you are using will present NO leakage path to ground.  Using several resistors in series is a good idea.  That said -- using a good HV probe is better yet.  I believe they are expected to work into a certain input resistance and will not be accurate when used otherwise.

Al

[WA1HZK]

If we can hear it, it's too much. If you have a decent filter cap of 30+ mf. and a 4+ H choke running under an amp on the final it probably won't be heard.
Keith
WA1HZK

[Opcom]

Maybe I'm paranoid, but if I were to measure HV ripple, and I didn't have a suitable probe, I'd build a 100:1 voltage divider. This would take a 3500v supply down to 35v. Then I'd either AC or DC couple the scope or DVM input. No way to zorch anything with this approach.

With my luck, I'f I hung a cap from the HV B+ to the scope input, even with a resistor to ground, and a bi-directional zener clamp, that cap would short out resulting in a massive flame-out of expensive equipment.

Jim
WD5JKO

The probe is best as suggested for safety reasons. 100:1 will let you easily see the hum if it is enough to ever hear over the air. I have an old B&K 40KV TV set probe designed to use with a 1M Ohm meter input and I use it frequently. I never handle it because it is old. I clip it in place then turn things on. Beware old probes because any invisible crack could let the sparks out. A good probe should have a ground lead and be made so that any short will be contained in the proble and hit the internal ground shield or whatever first instead of coming out of the handle.

But to help with a cap shorting, a HV fuse of a #30 wire or something can be put in series with it.
Or, a 1M resistor between the rig and cap, and a 1M res. from the other side of the cap to GND.
Or, use a really high voltage cap.

When I have done the direct measurement on a 3500V set, I used a 1uF 15KV oil cap. The diagram below shows a modulator test setup using a few resistors for either a voltage divider or not, depending. It would be OK for checking hum.

The set under test when that was done has a power supply with an 8uF input cap, a 4uH choke, and a 32uF output cap. The hum was hard to pick out from the "normal" noise from line variations.