Overload Board

[K9ACT]

My Class D is running pretty well now but I still blow FETs when I push it to 200W and much over 100% modulation so I have been just running it at 100W and not sure where to go from here.

I have the Overload Board and the bag of parts and I guess I can figure out how to hook it up but the instructions on the schematic assume a 45V carrier and I need the numbers for 30V. Also what is meant by 12V at input port in this setup?

One of the notes says that if it trips on normal voice peaks, nudge up the voltage cal a bit but this seems like inviting the problem I am now having.

I am thinking that I will just begin by shutting down the HV on overload but I need some advice on where best to do this.
As there is less current in the primary, this seems like the easiest place but it may not be the best for some reason I am not aware of. On the other hand, maybe it needs to be in the DC side.

Jack

[Gilly]

Hi Jack

Something must be wrong to blow FETs at the 200Watt level......all the time....
Is it only just one side that blows....?
How many go bang ! ..... 1 or 2......or the whole lot...!
Maybe use a current meter in each module and see if they draw appox the the same current...

I suppose you have checked the drive level at the Gates......its not over 20volts ? is it.......

Are your FET insulating washers new or used......

Could you post a picture of the TX...


Wayne

[K9ACT]

Hi Jack

Something must be wrong to blow FETs at the 200Watt level......all the time....
Is it only just one side that blows....?

How many go bang ! ..... 1 or 2......or the whole lot...!

It's pretty random but never more than one at a time. Well, on two occasions, it also took out the mod fet.

>Maybe use a current meter in each module and see if they draw appox the the same current...

Good idea.

>I suppose you have checked the drive level at the Gates......its not over 20volts ? is it.......

Not lately but it was never much over 12V when I did measure it.

>Are your FET insulating washers new or used......

Brand new.

>Could you post a picture of the TX...

This is not quite current but close enough...

Jack



Wayne

[Gilly]

Is it always the same FET in the same Module or is it random amongst them...

Are you using T200-2 cores in the filter....2 in total...the picture is a bit dark and could not see both cores....

OK too on the capacitors across the Drains in series....what type are they and do they get HOT at 200Watts.....

Also what goes faulty in the FET .... does the Drain to Source short or does the Gate pop....
If you still have the Blown Fets check them with a Multimeter.....

You would think if the Drain to source shorted then the Mod FET would go too making life unbearable to say the least.....

I enjoy testing a new TX...I'm always using my Dual Trace CRO checking the Gate and Drain Waveforms are much the same under varying conditions.....

If this was my TX ..... I would solve this problem before moving on to Steves protective board as it might get all too confusing...


Wayne

[kf1z]

My Class D is running pretty well now but I still blow FETs when I push it to 200W and much over 100% modulation so I have been just running it at 100W and not sure where to go from here.

I have the Overload Board and the bag of parts and I guess I can figure out how to hook it up but the instructions on the schematic assume a 45V carrier and I need the numbers for 30V. Also what is meant by 12V at input port in this setup?

One of the notes says that if it trips on normal voice peaks, nudge up the voltage cal a bit but this seems like inviting the problem I am now having.

I am thinking that I will just begin by shutting down the HV on overload but I need some advice on where best to do this.
As there is less current in the primary, this seems like the easiest place but it may not be the best for some reason I am not aware of. On the other hand, maybe it needs to be in the DC side.

Jack

Should kill the DC AFTER the filter caps, so that the HV DC is instantly removed from the circuit!

Don't worry about the 12 volts, don't worry about the 45 volt DC numbers... that is a rough set of numbers to start with.
So that you can key the transmitter to set these calibrations without the OL tripping.

What you need to do it set the voltages given, at the test points given, for your normal operating conditions.

[kf1z]

Jack,
Once you get the OL board assembled, and in circuit, we can help you get it set up correctly..

[steve_qix]

Killing the high voltage is not enough. The overload board, when PROPERLY connected to the PWM generator AND the power supply will instantly (within milliseconds) kill the pulse train, which stops the modulator output immediately. The high voltage is also shut down at the same time, but this takes a long (long in terms of semiconductors) time to happen.

The overload board will protect everything from overcurrent. If you have a parasitic and if that is causing devices to fail in the RF amplifier, the overload board will protect the modulator and other components from damage, but it can't do anything about the parasitics (IF this is what is causing the problem).

Regards,

Steve

[K9ACT]

The overload board will protect everything from overcurrent. If you have a parasitic and if that is causing devices to fail in the RF amplifier, the overload board will protect the modulator and other components from damage, but it can't do anything about the parasitics (IF this is what is causing the problem).

If I understand, it will prevent the drudgery of replacing FETs including the ones in the RF deck but I have to get rid of the parasitics to run it. (How do we teach this spell checker how to spell?)

If the problem is parasitics, it only happens at high levels of modulation and displays no other symptoms than blowing fets.

I am still waiting to hear from you on those high voltage Transient suppressors. I would at least like that base covered.
What do I need to do to get 10 of them?

I have everything but the semiconductors on the Overload Board but there is a heat sink in the package with no clue where it goes. It would fit the regulators but there is no hole in the board for the mounting tab. Was this an accidental inclusion or am I supposed to use it.

One more question, why is it necessary to shut down the PWM generator if the HV is shut down?

js

[K9ACT]

Is it always the same FET in the same Module or is it random amongst them...

Random with an occasional driver and/or mod fet.

>Are you using T200-2 cores in the filter....2 in total...the picture is a bit dark and could not see both cores....

Yes.. T200-2 They are mounted on an outboard plates so I can switch between 80 and 160. This pic shows one installed and the other in the background.



>OK too on the capacitors across the Drains in series....what type are they and do they get HOT at 200Watts.....

They are 10KV ceramics and don't get hot. Currently there is only one .001 instead of two in series. I am clueless as to the correct value. It does not seem to make any difference what the value is including none. I do not understand this business at all.

>Also what goes faulty in the FET .... does the Drain to Source short or does the Gate pop....
If you still have the Blown Fets check them with a Multimeter.....

I have checked them all and with one exception, it is always drain to source.

>You would think if the Drain to source shorted then the Mod FET would go too making life unbearable to say the least.....

This may be naive but I have a 5A fuse in the Variac which frequently fails when the drain shorts, perhaps saving the mod fet?

>I enjoy testing a new TX...I'm always using my Dual Trace CRO checking the Gate and Drain Waveforms are much the same under varying conditions.....

It's somewhat inconvenient to do this as I am set up because my 5 ohm dummy load isn't big enough for extensive bench testing and I can't get at anything when on the antenna or RF dummy but that's a dumb excuse.

>If this was my TX ..... I would solve this problem before moving on to Steves protective board as it might get all too confusing...

Sure but the board is a fallback bandaid that does not make me very happy either.

As said earlier, all it will do is save the fets but at least it might tell me exactly what level of modulation not to exceed.

Building it is more fun than changing fets.

Jack

[K9ACT]

The board is complete but I only tacked down the regulators waiting for someone to tell me what to do with the heatsink.

My next hang up is understanding what is meant by configuring the resets, hi, lo, assert, unassert, resistors that should or not be installed, etc. I installed them all before I noticed that note.

What is a good way to exercise or check out this board before installing it into the susytem?

js

[steve_qix]

The overload board will protect everything from overcurrent. If you have a parasitic and if that is causing devices to fail in the RF amplifier, the overload board will protect the modulator and other components from damage, but it can't do anything about the parasitics (IF this is what is causing the problem).

If I understand, it will prevent the drudgery of replacing FETs including the ones in the RF deck but I have to get rid of the parasitics to run it. (How do we teach this spell checker how to spell?)

If the problem is parasitics, it only happens at high levels of modulation and displays no other symptoms than blowing fets.

This is the typical parasitic problem - they usually show up only with higher voltages. It may NOT be a parasitic, but this is, as I say, the most common cause of blown components.

I am still waiting to hear from you on those high voltage Transient suppressors. I would at least like that base covered.
What do I need to do to get 10 of them?

Paypal me $1.50 each plus $3.00 mailing/packing No problem !

I have everything but the semiconductors on the Overload Board but there is a heat sink in the package with no clue where it goes. It would fit the regulators but there is no hole in the board for the mounting tab. Was this an accidental inclusion or am I supposed to use it.

The heat sink goes on the 5 volt regulator. This is documented somewhere, but I forget where right off..

One more question, why is it necessary to shut down the PWM generator if the HV is shut down?
js

Shutting down the power supply alone takes too much time. Relays are slow and they arc. By the time the power supply is off, the transistors are melted internally. The only way to stop the modulator output fast enough is to stop the pulse train. The pulse train stops MUCH faster than will the power supply. Of course, you also want to kill the power supply, but the pulse train is the first and most important line of defense.

[K9ACT]

The heat sink goes on the 5 volt regulator. This is documented somewhere, but I forget where right off..

I found it about 5 minutes before reading this. I knew I had seen a parts list somewhere but finding it again was a challenge.

> Of course, you also want to kill the power supply, but the pulse train is the first and most important line of defense.

So, what happens if I only kill the pulse train with the overload board?

I know you do not like my approach but just as a "hypothetical" question...

My power supply is always on, the PDM generator is always on and a breaker connects the mod output to the RF deck upon keying.

I could use one of the relay contacts to open the breaker or add another one in the HV line but in pursuit of simplicity, what would happen if I just shut down the PDM gen and then manually shut down the rest on overload?

More on paracitics later.

js

[K9ACT]

Preliminary checks of the Overload board do not look good.

TP1 reads between 3.5 and about 12v with or without 12v on the input.

The other test points read as follows:

TP2 2.5v
TP3 11.2v
TP4 2.5v

Overload Lo on Gen goes to Overload Lo on board.
When Reset Lo is grounded, pulse train starts and relay clicks.

That's about as far as I can get with it.

Jack

[kf1z]

If you use "reset low", then you should have to put +5v on it to start pulse-train etc...


>"TP1 reads between 3.5 and about 12v with or without 12v on the input."


Do you mean the voltage is jumping around?

[K9ACT]

If you use "reset low", then you should have to put +5v on it to start pulse-train etc...

That's what I expected but it's not working that way.

>"TP1 reads between 3.5 and about 12v with or without 12v on the input."

>Do you mean the voltage is jumping around?

No. That is the range I see when I turn the pot. The low end is around 3 and after about a quarter turn it jumps to about 11. The range is somewhat different depending on whether the 12v is applied but it never goes below 3 and is supposed to be set at .8 for starters.

I don't understand the reset at all. I would expect to push a button after an overload to reset it but it appears that the reset condition must be maintained and this doesn't make any sense to me.

I also find assert and unassert rather confusing terms. And then the unused reset is to be "de-asserted".

I am lost but I guess I need to get the board working first.

js