UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > General Vintage Technology > Components and Circuits

Notices

Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets.

Closed Thread
 
Thread Tools
Old 2nd Feb 2017, 11:00 pm   #1
Wendymott
Octode
 
Wendymott's Avatar
 
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,782
Default JFET SSB Linear amp problem

Hi peeps. Further to my question regarding Torroids, I took the plunge and made various versions until I found values that suited my circuit. "see attached". This worked reasonably well as a lashup...... however when I committed to copper, I had a few problems.. like self oscillation, which appeared to be caused by the bias to the fets Q3 and Q4.
Now my understanding is that as they are "single ended" amps, they should operate in class "A", but I dont know how to work out when each device is operating in that mode. What drain current should I set each fet to ?
I also tried damping both Q3 and Q4 torroids with resistors, but this had no effect......
To re itterate... this is a 40 metre transceiver, but will be upgraded to most bands if successful.
I only need about 5W PEP from a 25mV drive, as I intend to drive an "outboard" linear to 100W if required.
The obvious solution... if I cannot fix this.. is to "crib" the larger PA circuit and under drive it.
Your thoughts as usual would be most welcome
Attached Thumbnails
Click image for larger version

Name:	PA Module.jpg
Views:	174
Size:	72.0 KB
ID:	136885   Click image for larger version

Name:	Linear PA ver1.jpg
Views:	342
Size:	42.3 KB
ID:	136886  
__________________
Should get out more.

Regards
Wendy G8BZY

Last edited by Wendymott; 2nd Feb 2017 at 11:01 pm. Reason: Added pictures
Wendymott is offline  
Old 3rd Feb 2017, 12:46 am   #2
Terry_VK5TM
Nonode
 
Terry_VK5TM's Avatar
 
Join Date: Oct 2010
Location: Tintinara, South Australia, Australia
Posts: 2,324
Default Re: JFET SSB Linear amp problem

The two heatshrink? covered coils on the lefthand side, assuming they are either L2 & L3 or L3 & L4 look to be too close together and could be coupling to each other.

One should be at right angles to the other to reduce the coupling.

Also, the 10 ohm gate resistors should as close as is physically possible to the gate leads to the point of being directly soldered to the leg of the FET.
One possibility is to remove the through hole ones, link the pads then cut the track right at the gate lead pad and use smd resistors to bridge the gap.

Terry VK5TM
Terry_VK5TM is offline  
Old 3rd Feb 2017, 2:52 am   #3
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: JFET SSB Linear amp problem

I think that the coils with heatshrink are probably L5 and L6 in the LPF?

Note that the thread title should really be "MOSFET SSB Linear amp problem" and the symbol is wrong for the IRF510 as it should be a MOSFET and not a JFET. But this isn't a big deal I guess.

There's a connection error in the schematic because C15 should go to the drain of the IRF510 Q4. Also, the whole Tx chain looks a bit scary in terms of how the stages are designed and connected. Maybe you can get this thing stable but it doesn't look good to me on paper at least.

The other point to note is that I can't see how you can get more than maybe 2W from it when running with that LPF and a 12V supply with a 50 ohm antenna load.

Note also that a 5th order LPF made with a pair of identical solenoids and leaded ceramic caps like this will be prone to re-entrant modes up at UHF where the stopband can have big upward spikes in the response. You could easily get a couple of narrow/spiky bandpass responses up around 500MHz where the loss could be just 20dB or even worse. You can fix this a number of ways, either with a roofing filter or make the two inductors subtly different in terms of L/D ratio but still getting 1.4uH. This can 'spread' the spiky BPF responses in the stopband and make a useful improvement in filtering performance
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 3rd Feb 2017, 3:05 am   #4
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: JFET SSB Linear amp problem

Quote:
Originally Posted by Wendymott View Post
Now my understanding is that as they are "single ended" amps, they should operate in class "A", but I dont know how to work out when each device is operating in that mode. What drain current should I set each fet to ?
I've not used the IRF5xx series very much but you can get some idea of linearity by looking on a datasheet that shows the curves for Id vs Vgs voltage vs Vds. Normally you would try and locate a sweet spot where it is most linear. My limited experience of these devices has shown me that you generally have to bias them quite hard to get them linear enough for SSB so you may end up having to bias it at several hundred mA. I think some CB users bias the IRF5xx devices up above 0.5A to get decent linearity on 27MHz.

I would much rather make a PA using typical BJT transistors used in SSB CBs. At 7MHz you would need to use some negative feedback to tame the device gain but a ~5W BJT amp can be run at a bias of just 50mA for SSB use. Much better efficiency especially for low duty stuff like SSB. So the heatsink will run much cooler on a long over.
__________________
Regards, Jeremy G0HZU

Last edited by G0HZU_JMR; 3rd Feb 2017 at 3:20 am.
G0HZU_JMR is offline  
Old 3rd Feb 2017, 10:03 am   #5
ionburn
Heptode
 
ionburn's Avatar
 
Join Date: Dec 2014
Location: Lincoln, Lincolnshire, UK.
Posts: 583
Default Re: JFET SSB Linear amp problem

Quote:
Originally Posted by G0HZU_JMR View Post

I would much rather make a PA using typical BJT transistors used in SSB CBs. At 7MHz you would need to use some negative feedback to tame the device gain but a ~5W BJT amp can be run at a bias of just 50mA for SSB use. Much better efficiency especially for low duty stuff like SSB. So the heatsink will run much cooler on a long over.
The issue with this approach these days is that the BJT transistors commonly used in the output stages of CB's are in short supply and commonly faked. There is a trend to replace faulty output stages in CB radio's with FETs
ionburn is offline  
Old 3rd Feb 2017, 10:23 am   #6
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: JFET SSB Linear amp problem

There are still real, unfake TO5 RF transistors around, good for a couple of watts each. A push-pull amp with a pair on each side with a bit of emitter ballasting resistance would make a useful driver. The problem is these things have gain to about 1GHz so VHF-savvy construction is a must. Tens of thousands of genuine 2N3866 were distributed for free around the QRP groups. There must be some still kicking around.

SSB can be amplified in a class A-B stage. Being cut-off for almost a half-cycle creates energy at harmonics of the RF frequency, but provided action is linear during the on-time, there won't be horrible distortion of the audio baseband being conveyed. It's counter-intuitive, but it works. The RF harmonic filter does the job of removing the distortion products. Square law devices also help by creating chiefly even-order products at frequencies far from the intended output, while making relatively little odd-order stuff which can turn up inside and surrounding the wanted modulation template.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 3rd Feb 2017, 11:30 am   #7
Wendymott
Octode
 
Wendymott's Avatar
 
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,782
Default Re: JFET SSB Linear amp problem

I knew I could rely on you... Terry the two inductors are the LPF L5/L6 and I take the note from Jeremy about making them.. Non identical... I also thank you all for the other information.....and I will start again, with all the comments made...I will let you know the outcome.
Just toshow this is not a "pie in the sky" project I attach photo's of the near completed item.
Attached Thumbnails
Click image for larger version

Name:	6 Front final view.jpg
Views:	135
Size:	68.3 KB
ID:	136908   Click image for larger version

Name:	10 Internal view topside.jpg
Views:	142
Size:	73.0 KB
ID:	136909  
__________________
Should get out more.

Regards
Wendy G8BZY
Wendymott is offline  
Old 22nd Feb 2017, 9:13 pm   #8
Wendymott
Octode
 
Wendymott's Avatar
 
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,782
Default Re: JFET SSB Linear amp problem

Hi peeps. An update of the final construction of the PA. I ditched the 2 X Fet Class "A" series circuit and adopted a 2 X FET Push pull output stage with 30 m/A bias per device. The overall output is 10W PEP into a 50 R 100W load. The lashup worked well, but obviously could not be allowed in the final version. The pcb was designed and lo and behold a 10W 7 Mhz Power oscillator was the result. After a few mods were made with no effect, I looked at the pcb copper and made some cuts. Voila...however when the base plate, which doubled as the heatsink was added, again it oscillated like a "goodun".
The remedy was to use an isolated pillar at the RF input corner, with the other 3 pillars grounded.
The amp was thoroughly tested, with a bench PSU. Time to fit into the project case.
It was now obvious that the internal power transformer was lacking in capacity. If you read my other thread ref Danbury transformers, I eventually made a suitable power supply, supplying 14V @ 2.2A and 12V @ 500 m/A for the low power sections.
Finally "airtested" today with 5 +7/9 into Banff Scotland and an EI in Donnegal .
I will post the full circuit data in the "Homebuilt" section.
Please see the schematic and RF output photo.
Attached Thumbnails
Click image for larger version

Name:	2002A.jpg
Views:	88
Size:	53.7 KB
ID:	138045   Click image for larger version

Name:	2002B.jpg
Views:	142
Size:	53.1 KB
ID:	138046   Click image for larger version

Name:	2002G.jpg
Views:	112
Size:	73.1 KB
ID:	138047   Click image for larger version

Name:	2002II.jpg
Views:	88
Size:	60.2 KB
ID:	138048   Click image for larger version

Name:	Linear PA 5 B.jpg
Views:	299
Size:	41.3 KB
ID:	138049  

__________________
Should get out more.

Regards
Wendy G8BZY
Wendymott is offline  
Old 23rd Feb 2017, 2:12 am   #9
Terry_VK5TM
Nonode
 
Terry_VK5TM's Avatar
 
Join Date: Oct 2010
Location: Tintinara, South Australia, Australia
Posts: 2,324
Default Re: JFET SSB Linear amp problem

Good show Wendy.

Yes, RF can be a real 'black art' at times.

Terry
Terry_VK5TM is offline  
Old 25th Feb 2017, 12:25 am   #10
VHF_BUILDER
Diode
 
Join Date: Feb 2017
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 9
Default Re: JFET SSB Linear amp problem

Good work Wendy, glad to see you are making good progress - much of my radio project is still in Eagle CAD and on breadboards

Just so you know, I have chosen a class AB push pull 2N3866 pre-driver circuit which will drive a pair of RD15HVF1 mosfets in class AB and finally a pair of RD70HVF1 in the same configuration for the final PA.

Just out of curiosity what technology did you use for the VFO?

Cheers

Rob.
VHF_BUILDER is offline  
Old 27th Feb 2017, 12:18 am   #11
Wendymott
Octode
 
Wendymott's Avatar
 
Join Date: Jun 2013
Location: Huddersfield, West Yorkshire, UK.
Posts: 1,782
Default Re: JFET SSB Linear amp problem

Hi Rob.... Sorry not seen this till now....I was thinking the same pre drivers , even bought some 3866's....I have started the Linear PA PSU... just finished laying out the pcb for the capacitor bank 5 X 10,000 ufs.. The Linear module was bought from the Inscrutible Chinese, except they sent the 100W version instead of the 50W... Result..
My Main VFO is a AD9850,DDS with software from my associate in Aussie... if you are interested I can Pre programme a 16F328, but I cannot give you the code. If you look in the Components and Circuits section the schematics are there.
The DDS gives me a 1 Hz resolution at 7 Mhz.
Thanks for the interest
__________________
Should get out more.

Regards
Wendy G8BZY
Wendymott is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 7:08 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.