The AM Forum
March 28, 2024, 06:43:04 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: 1 ... 11 [12] 13   Go Down
  Print  
Author Topic: The 6M-160M HPSDR Transceiver Project Begins  (Read 222705 times)
0 Members and 2 Guests are viewing this topic.
KF1Z
Member

Offline Offline

Posts: 1796


Are FETs supposed to glow like that?


« Reply #275 on: February 15, 2011, 08:12:42 PM »

I have an Erbtec board here I never used... scavenged some parts, that's about it...

Tom,  do you want the 150s off it?
Logged

K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #276 on: February 15, 2011, 08:45:26 PM »

Mike,

Here's the info write-up on Hermes:

http://openhpsdr.org/wiki/index.php?title=HERMES

Yes, it does have built in comm that does USB to computer. No need for Ozy.  Frank, is there any reason to include Metis with Hermes?

Looking over the specs, eight independent receivers using the same antenna, and all the other nice stuff, it's quite a rig.


Bruce, so you have an ErbTec too?  Well, since I've lost the MRF-150's, I could use a new pair when Frank finalizes the mods. From the thread you probably see I'm trying to get it cleaner. It's the cores in the driver, Frank thinks.  Send me a price for the two MRF-150's and I'll pop them in once the design is finalized. Tnx, OM.

T

Hermes: On the air and working well. Just waiting for it to become available to public.


* Hermes 6M-160M Transceiver.png (431.71 KB, 500x375 - viewed 1041 times.)
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #277 on: February 15, 2011, 10:05:05 PM »

The Hermes design was modified to the metis Ethernet interface. This is a stand alone transceiver.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #278 on: February 16, 2011, 11:11:48 PM »

Finally some success.  The HPSDR transmitter system is working and very clean.

After a week+ of playing with the TMC amplifier as a linear for the HPSDR rig, I decided it was just too much voltage, blower noise and space for a measly 200w pep.  Instead, I went back to the FT-102,  using it as a linear amplifier.

I increased the RF negative feedback cap from 6pf to 18 pf. This really made a difference in its cleanliness, but was harder to drive, of course. Also idled up the three 6146's to 110ma.  I tapped into the pre-pre driver of Q9's base and the Penny's 10mW will drive the pants off it there.

The result was 50w pep output that is close to -50db 3rd! VERY clean side products. If I push it harder, I can get out 150w pep but at about -40db 3rd.  And, the FT-102 is self contained, small and quiet. A no brainer. What a FB 10mW > 150W linear it makes.

I tried driving the 3-500Z's with this system and it worked FB with just ~30W pep drive.  My Mercury RX is still out for repair, but using a separate RX, I plan to get on the air with this system for some tests ASAP.  

Frank, I got rid of that ant T/R keying relay in the HPSDR and made a direct connection to the FT-102 linear input from Penny..  The adjustable pad works great.

Rob/AEX , you around tmw?

T

Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
W1AEX
Un-smug-a-licious
Contributing
Member
*
Offline Offline

Posts: 1488


Apache Labs SDR


WWW
« Reply #279 on: February 17, 2011, 11:10:21 AM »

Finally some success.  The HPSDR transmitter system is working and very clean.

Rob/AEX , you around tmw?

T

Hi Tom,

It looks like you have been busy and making lotsa progress! Shoot me an email or a PM and let me know when you plan to put your newest creation on the air. I'll have my staff rearrange my busy schedule to make time for Mr. Vu today.

Rob
Logged

One thing I'm certain of is that there is too much certainty in the world.
KF1Z
Member

Offline Offline

Posts: 1796


Are FETs supposed to glow like that?


« Reply #280 on: February 17, 2011, 11:27:28 AM »

I'm wading through... trying to find the best place to look for information/announcements when Hermes will become available..

The last updates I found were from June 2010.... ?

Logged

K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #281 on: February 17, 2011, 11:50:04 AM »

Sounds good, Rob, thanks!

I'll send you a PM later in the afternoon.  Otherwise leave your RX on 3765 LSB about 3PM.


Bruce, take a look at the archives for any talk about Hermes, otherwise, join the reflector and axe them directly what is going on with Hermes.  I would like to know and wud axe, but have spent my nickel already about the keying spike and all the help I received getting up and running... Grin

Right now there is a lot of thread traffic about the new comm board, Metis that was just shipped. Seems to be working FB.

http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/

http://openhpsdr.org/

As you know, this is not a commercial company, so thangs go slowly and there are often shortages and delays. But the end product is worth waiting for.  

BTW, Please let me know the price for those two ErbTec MRF-150's. I'd like repair it for some new mods. Thanks.


T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Mike/W8BAC
Contributing
Member
*
Offline Offline

Posts: 1042



WWW
« Reply #282 on: February 18, 2011, 03:59:33 PM »

Quote
Looking over the specs, eight independent receivers using the same antenna, and all the other nice stuff, it's quite a rig.

Seems odd Tom that they would have eight independent receivers and not dual diversity. I wonder if the receivers are fully independent and can operate on different bands and modes at the same time?

I'm just starting to read up on the Hermes project. By the time they ship I'll be up to speed.
Logged
KF1Z
Member

Offline Offline

Posts: 1796


Are FETs supposed to glow like that?


« Reply #283 on: February 18, 2011, 04:26:37 PM »

I think Alberto was the last to even ask about it..

quite a while ago....

There is likely so much interest, that those of us outside the circle will not see one for a very long time.  Wink
Logged

K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #284 on: February 18, 2011, 04:42:13 PM »

Mike,

From my basic reading,  "independent receivers" means it's one hardware receiver, one antenna source, but being broad-banded as it is, they can write software to select the particular frequencies to listen to.  Maybe Frank can expound on this.

In contrast, dual diversity requires two truly separate sets of receiver hardware because of two antenna sources.
The tech reason for problems using two Hermes in diversity were related to synchronizing them  and other thangs that I did not understand fully. They said it COULD be done but wud be too complex for whatever reasons.

I understand that they CAN operate on different bands at the same time, but the frequency difference is limited to some maximum. It's in the specs somewhere. 


Bruce,

Check to see if there is a waiting list for Hermes. There are for other boards usually.  They will make as many boards that are ordered for the first run, so you have just as much chance as anyone to get one, I wud think. I agree they shud be a hot item. Heck, they even built in a 2.5 w audio amp to drive speakers directly along with a built in 75M Yagi in software..  Grin


T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Mike/W8BAC
Contributing
Member
*
Offline Offline

Posts: 1042



WWW
« Reply #285 on: February 18, 2011, 05:15:47 PM »

Quote
one hardware receiver, one antenna source, but being broad-banded as it is, they can write software to select the particular frequencies to listen to.

That is about what I expected. One mode and up to 8 locations. I'm still wondering if it (the receiver) is truly broad banded. I read in the Wiki information Hermes has "Uninterrupted coverage from 50kHz to 54MHz". It would be beyond cool if you could set your receivers to hot spots on say 40, 80 and 160 and work 10. Limitations being the antenna system.

When you mentioned a second discreet receiver and antenna port, the light came on. Some things stay the same even with the magic of software.

What do you know about the 15 watt Pennywhistle PA? Will that be clean enough to drive my single 8874 to complete the driver stage?
Logged
KF1Z
Member

Offline Offline

Posts: 1796


Are FETs supposed to glow like that?


« Reply #286 on: February 18, 2011, 05:18:56 PM »

Tom,

Already looked for a list, doesn't appear to be one...

Not the easiest set of webpages to navigate!  Roll Eyes

I'll send you those 150s... If they are good, send me some $., if not.. don't.!

I have no good way to test them. They are still on the board.

Logged

K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #287 on: February 18, 2011, 05:35:07 PM »

Thanks Bruce.  Maybe the Erb amp can ride again.

Mike,

The PennyWhistle 16W linear is rated about -30db 3rd below tone or -36db below pep.  Personally, I wud look for something cleaner since the Penny exciter is -55db ++. What a waste of a clean exciter.   Get something at least -40db if possible. 

Another thing, to drive your amplifier, you would be pushing the PennyWhistle to full output. From the tests I've run, most amps are cleaner when backed off and (tube amps) loaded heavily. My SS FT-1000D acts better at lower levels too.  So find a 100w -40db 3rd IMD linear amp and work it easy at 20-30W. That's what I'm trying to do.

Wish there was something out there that did 10mW in > 100w out, 40V devices, -50db 3rd, right off the shelf cheap. There's lots of nice SS amps on eBay, but they're VHF, UHF, etc. There's not a big market these days for HF amps.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Mike/W8BAC
Contributing
Member
*
Offline Offline

Posts: 1042



WWW
« Reply #288 on: February 18, 2011, 05:51:58 PM »

Tom,

I need a solid 100 watt capable driver for the 4CX3000 amp (Gates Vanguard 1). I posted here some time back regarding the use of my Alpha 374 amp with one of the tubes pulled. It dose a good job at 100 watts but I'll need about 15 watts predrive and Pennywhistle is tempting. I'll have to look at the Motorola rf handbook to see if they have any magic devices with specs like -40db 3rd IMD but I'll bet it will take two stages at least. Every stage adding noise.
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #289 on: February 18, 2011, 06:03:28 PM »

I have my erb amp doing -40 at 10 watts out but have more work to do with the driver transformer and feedback. A pair of MRF150s class A at 6 amps bias are rated by Motorola at -50 dBc at 100 watts out.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #290 on: February 18, 2011, 06:31:47 PM »

One might axe, if a final linear amp is only -30 db 3rd, why bother with a super clean exciter and IPA? There's a few reasons.

1)  A -30db final linear will always be dirtier than -30db, depending on the driver. If the driver(s) is very clean, then the final will be close to -30db.  If the driver is  also -30db, then the final will be down around -29db.  If the driver is -27db, then the final will be below -27db. You get the idea.

2)  If the driver is very clean, then why not improve the final linear by idling it higher, loading it more heavily or running it at less power output?   If the driver is marginal, then cleaning up the final will have much less effect.

3) When operating and power is needed for say breaking a pile up, pour it on. But when the QSO starts, back the power down. If the driver is clean, the final will clean up more at reduced power. When we are transmitting alone, people notice our splatter. In a pileup or on a very noisy band, we blend in.


Mike, that Alpha 374 could make an excellent IPA driver for the 3X3 linear.  Load it very heavily (lots of C2 capacitance giving a broad plate dip) at low power and see about adding some higher plate current idle by reducing the cathode zenor.  Maybe reduce plate voltage with the xfmr tap.   I did that with a Henry 3-500Z amp and it's stellar clean.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #291 on: February 19, 2011, 10:45:15 AM »

According to HPSDR software writer Doug/W5WC, the unkey spike problem has been fixed and is now available in the latest PowerSDR software:  (sent out on the reflector today)


"Updates were added to PowerSDR/OpenHPSDR v.1.19.3.3 that addresses the issues users are having with “hot switching”. A noticeable 80mS or more of rf was still present after returning to receive. The delay is not consistent and varied greatly depending on the computer in use. This may not correct the problem for everyone so if you find my default settings or limits need adjusting please let me know.

 Added to the Setup/Options tab are three controls that to help fine tune your radio.  Enable PTT Delay will enable the PTT Delay up/down control. The PTT Delay control adds delay between the release of the PTT until the PTT Out line drops. The third is the MOX Delay up/down control. This allows the user to fine tune the delay between when the buffers empty and the dropping of the PTT out line.  If this control is set to low the carrier will still be present at the output after returning to receive.  I added the MOX Delay because of the vast differences noticed with latency in systems I used to test with.

The PTT Delay is not active for the CW modes. Delay control for cw break-in is already present. The MOX Delay is active for all modes.

Also added was a display distortion fix sent to me by Joe, K5SO. Many thanks for that info.


73, Doug W5WC"
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #292 on: February 19, 2011, 10:48:06 AM »

Here's the new Hermes transceiver status as reported today. Looks like they had a working version months ago but decided to add some further improvements:



"Hermes beta with PHY is under development, Phil has the FPGA code almost ready and we are now in the process to try and see if we can add a 2M front end for Hermes without degrading HF performance or adding spurs.

The delay has been on my end i'm afraid, morning QTH has been rather busy lately.

I have completed the placement for Hermes with ONLY the HF components so far, a lot of progress has been made on the PCB and we are confident that we'll have the final version ready very soon.

73s,

Tony"
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Mike/W8BAC
Contributing
Member
*
Offline Offline

Posts: 1042



WWW
« Reply #293 on: February 20, 2011, 09:01:33 PM »

I'm curious Tom, I'm sure you have a reason for wanting 100 watt IPA that runs on 28 volts instead of 13.8 and I understand the distortion issues but if somebody was entertaining the idea of developing such an amp wouldn't it be one that uses the same voltage as Hermes?

Mike
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #294 on: February 20, 2011, 09:05:44 PM »

Higher volatge amplifiers are cleaner. Hermes uses high frequency OP amps as output stages.
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #295 on: February 20, 2011, 09:29:15 PM »

I had some thoughts on the Erb amp and cleaning it up.
Tom and I both agree the drive signal to the gates of the final is the first signal that gets distorted as you increase power. The 300 watt Erb amp is loaded with 2 10 ohm resistors in parallel on each final gate. I noticed that my 1200 watt Erb amp is loaded at a higher resistance. 24 ohms if I remember or maybe 2 in parallel. Both are the same basic design using MRF150s.
So if the input resistance to the final is increased by a factor of two a couple things will happen. First it will be easier to drive the final with a clean waveform. Second the higher input resistance will increase the negative feedback. Both the 300 watt and the 1200 watt amps use the same value feedback resistor.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #296 on: February 20, 2011, 09:30:37 PM »

Mike,

The HPSDR boards require very small power (+12V, 5V) to run them. Any external linear amp would need a big supply in comparison and wud be separate anyway.  Mobile operation wud be an exception, but I'll bet 99% of all are home stations.  Actually the newer Fets are 50V or so and the specs for linearity are MUCH better than the 13V devices, as Frank said. It's very worthwhile to get something that is at least -40db 3rd, in my opinion.

BTW, I have the SS linear problem solved. A good friend is swapping me the PA unit and LP filter unit from a junker FT-1000D in exchange for some gear I don't use anymore. I have my own FT-1000D and at 40W it is about -40db 3rd. This shud work out fine for the HPSDR as a 40W driver or a barefoot stand-alone 200W out pep rig.

Frank, I wonder if Penny will accept ALC back from the "protection board" of the 1000D PA Unit?  It generates a signal when swr is high or power exceeds a certain threshold.  Or maybe I wud be better off to trigger a fast SS switch at the input of the RF drive to the 1000D PA unit?  One way or another, I don't want to run that thing naked.. :-) It requires 30V for the finals and uses MRF-422's, 150W bipolar devices rated at -30db 3rd at full power, but my tests show much better specs at reduced power.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #297 on: February 21, 2011, 12:27:56 PM »

Here's an interesting post on the HPSDR reflector. It was a reply to my search for a clean SS linear. He's water cooling a pair of MRF-429's, a CCI module with a CPU water cooled assembly, idling at 50V, 8A = 400W. He sees -50db 3rd at 100W out.

This technique requires lots of heat dissipation for sure.

Frank, I tried both Erb and the CCI MRF-150's at low power, class A levels for short periods and didn't have such good IMD numbers.  Maybe I can find a way to water cool the junker 1000D PA module as he suggests.

T


------------

Tom,
I am using behind Penelope a CCI kit with two MRF433 as a driver stage to maximal 8 W in class A (bias current 3 A) with IMD3 better than -50 dB overall at 7 MHz.
I am just measuring in detail a next stage following with two MRF429 (50 V), that's a CCI kit as well (AN-758).
If run in class A with 8 A quiescent current it delivers up to 100 W with
IMD3 a little less than 50 dB below one of the test tones (clearly better than 50 dB at your envisaged 30 W).

Since 400 W of steady state power must be dissipated in this mode of operation I have water cooled the transistors by using a computer processor cooling system.

I am just busy measuring all specifications in detail, that's IMD3 as function of power at all bands.
I will send a paper on it to Stan for publication in the next PSR, the TAPR quarterly newsletter.

Best 73, Hans, DL2MDQ.
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #298 on: February 21, 2011, 01:44:50 PM »

I'll take MRF150s over MRF429 or MRF433s any day. IMD5 and above better with FETs. I have a pile of MRF429s. Bias at 8 amps man that is making heat.
I don't know of any ALC back to penny check the manual.
I'm headed to the shack to integrate metis. I hope to be on 75 later.
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #299 on: February 21, 2011, 02:02:22 PM »

I might chop out my single board CCI amp using four MRF-150's and see what it does for IMD. It has the combiner on it and is two boards presently. (eight FETS) Maybe the combiner and splitter was adding problems. It was no better than the ErbTech. I heard an ssb guy on with four of these CCI boards a year ago and he was 10kc up the band. I'm probably just wasting my time with that particular design.


The 1000D PA unit sends the ALC back to the IF board. Maybe I can trigger a SS device at the PA Unit RF input to open up when there is a fault threshold.   Even an active  pad attenuator would fold back the RF more gradually.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Pages: 1 ... 11 [12] 13   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.078 seconds with 19 queries.