Hello Band Volt mod aficionados

[Mark Herbert]

Hi has anyone done the modification for the Band Volt output for band switching a PA?

I am in the process of doing this mod but the pictures on github added by Alex

https://github.com/softerhardware/Hermes-Lite2/wiki/Band-Volts

are not clear as to where the wiring goes e.g. for ground and what are some of the extra connections to the DB9.

If someone could add a bit more detail then I think I can get this up and running.

Thanks

Mark G1LRO

[Reid Campbell]

Hi Mark,

I produced the design for the band volts. I think Alex must have corrected my bad spelling or grammar :-)

I was a bit naught with the ground and picked it up from the chassis screws holding the back panel PCB. As for the extra wire, that is the PTT signal. I wanted it to be on the 9 pin as well so I could get away with one lead to my XPA125B amp.

Let me know if you have any other questions.

Cheers

Reid
Gi8TME/Mi0BOT 

--
You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/78a72933-ab8f-4b6b-a24e-7209a9fec8f8n%40googlegroups.com.

[Mark Herbert]

Hi Reid

Thanks for the helpful response. I had wondered if the ground was coming through the chassis, I'm happy to go with that since it's tried and tested :)

Regarding the PTT signal, I was thinking about handling this by having a flying phono plug coming out of the DB9 shell, but picking it up internally seems like a more elegant solution. Did you connect to the back of the phono socket on the output board or is there a more suitable point to pick this up from? 

Thanks for developing the mod, I'm looking forward to connecting up my XPA125B amp for automated operations too.

Regards

Mark

G1LRO

[Reid Campbell]

I picked up the EXTTR on pin 2 of DB13. This is a 2x5 header and pin 2 is on the second row opposite pin 1 which is marked as a square pad. The row that contains pin 1 is masked by the header that joins the HL and the N2ADR boards together.

Cheers

Reid
Gi8TME/Mi0BOT

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/20f1795a-7c3f-4499-8997-bcca74c7aff1n%40googlegroups.com.

[Mark Herbert]

Hi Reid

I've gone ahead with the mod as prescribed, but I seem to have a problem with all the band voltages being lifted by about 120mv

Do you have any ideas for how I could bring this down across the board? 

I've tried adding a load to the output and that fixes the lower voltages, but then affects the higher ones.

Any thoughts?

Kind regards, 73

Mark

G1LRO

[Reid Campbell]

Hi Mark,

I've had a quick look at this and I think the ON pulse is being stretched by the M1 MOSFET. Try taking R1 down to 2K and let me know how that goes.

Cheers

Reid
Gi8TME/Mi0BOT

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/b7475f3d-1746-4937-8ca0-0190634411d4n%40googlegroups.com.

[Mark Herbert]

Hi Reid

Genius!

That completely resolves the issue:

Many thanks for your help.

During my efforts to resolve this I did notice that the input to the XPA125B is very high impedance and the calibration of the PWM is relying on there being no load. Perhaps if this goes into more formal release, then having the output loaded to say 100k and the calibrating the PWM will remove the sensitivity of a linear amp with a lower impedance band volt input which will have an effect on the RC network.

Anyway, super stuff, your help on this much appreciated.

73 G1LRO 

[Reid Campbell]

Great, now I have to figure out why mines working OK. Are you using the specified FET?

I'll up date the Wiki to reduce the resistor value when I get time.

Cheers
Gi8TME/Mi0BOT

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/49fb7126-1cec-4455-bb15-33acf89a97e1n%40googlegroups.com.

[Mark Herbert]

I did have to substitute the FET as the one on the BOM was unobtanium at the time. the replacement I used was a PMV50UPE

https://www.digikey.co.uk/en/products/detail/nexperia-usa-inc/PMV50UPE-215/3679494 

This was listed as a compatible part, but may explain the difference in behaviour.

I have a spare two here if you want me to pop them in the postbox for you to your QTH on QRZ?

Cheers

Mark

G1LRO

[Clifford Heath]

> I did have to substitute the FET as the one on the BOM was unobtanium at the time. the replacement I used was a PMV50UPE

> This was listed as a compatible part, but may explain the difference in behaviour.

Looking at the data sheets to compare the original IRLML6402 with the substitute PMV50UPE,
I first checked the Total Gate Charge (increased from 8nC to 10.5nC, not a big shift),
then the turn-off delay time (increased from 588ns to 2180ns - significant),
and lastly at the drain resistance as a function of Vgs - Bingo!

The original turns off much earlier as Vgs is drained away by R1. The PMV50UPE Vgs needs to drop to under a volt compared to the original at a couple of volts.

MOSFET gates are quite capacitive, and when draining the charge using just a resistor, you get an exponential decay to the knee in the curve where the output turns off, so both the gate capacitance and the knew voltage affect the turn-off time.

Clifford Heath.

[Mark Herbert]

Thanks for digging in and getting to the bottom of this. A very helpful explanation of the root cause of my problem

On the positive side, there are two Mosfet configurations that can be used now; useful if the  IRLML6402 remains hard to get.

I'm happy to send a few of the replacement devices to Reid so he can validate with his unit that's acting up.

All the best

Mark 

[Reid Campbell]

Hi Mark, Clifford,

Clifford, I think you are spot on with the Vgs, that would have more of an affect than the turn off delay.

Mark, the interface has been working OK for me but I might test the 2K resistor in the future. There is no need to send one of your PMV50's, keep it for you next HL2 :-). Also thanks for building the interface and highlighting the issue with different devices.

Cheers

Reid
Gi8TME/Mi0BOT

--

You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/5ea632d6-b255-4cb1-a86d-954f87cef4c7n%40googlegroups.com.