Moon-net August 2016

moon-net@mailman.pe1itr.com

164 participants
190 discussions

Start a nNew thread

Re: [Moon-Net] UPDATE - Drive, bias etc. levels for K2RIW 432 parallel 4CX250B amplifier.

by w8zn54＠verizon.net

6 years, 9 months

6cm AW G3LTF results

by Peter Blair

I was only able to be active on part of Saturday and I worked SV3AAF, HB9SV, II3EME ( 3rd band!) SM6CKU, OK1DFC #58, UN6PD #59, OK1KIR, SQ6OPG, UA3PTW, SP6PGP and OZ1LPR. Also heard UA3TCF.All QSOs on CW. Thanks to all for the nice contacts, I sorry I wasnt able to be there for the NA stations but we had a big family gathering this weekend. 73 Peter G3LTF

6 years, 9 months

6cm Sunday

by SM6CKU

Hello again, Another five EME QSO's today plus one rain scatter QSO while pointing the antenna at the moon. That was SM6PGP calling and I could hear his RS signal on my echo, but I could also see his moon echo some 9 KHz away. Normally we do not hear each other when pointing antennas to the moon. The EMEers worked were OH2DG, SQ6OPG, WA9FWD, K2UYH and VE6TA. Overall weekend result being 15 stations worked incl. 8 initials. Thanks everybody! 73 de Ben - SM6CKU 4m dish and 45 watts

6 years, 9 months

Re: [Moon-Net] BF188XR PCB

by Bent Andersen

Hi Gabi I use 2x BLF188xr at 144mhz and the PCB is from W6PQL, and the system is water cooled.. look more at http://oz1ct.dk/konstruktioner/144mhz-sspa-2xblf188xr-water-cooled.html 73" OZ1CT Bent

6 years, 9 months

Re: [Moon-Net] 6m EME Almost QRV

by Gedas

Hi Ed. I hope I did not mislead you about the transmitting caps. it's not that they did not work....they worked fine and could handle high heat. The issue was their values were not real close to what they stamped as, even taking into acct their tolerance. All of the caps I tried worked fine and provided me with a nice smooth Bode plot but it was not until I hand picked the precise values I needed (per the design) did I see about a 10 dB improvement in performance. I guess if you measure ur SSPA and the harmonics and spurs are not as bad as mine then probably any caps will work but in my case I had to make sure each cap value was very close to what my designed called for. Oh, I also saw that design you are building up.....it sure looks neat. The only reason I did not go that route was because I already had this brass cased Harris kW LPF that started out life at 30 MHz so it was easier for me to change the topology a little bit along with some different values and ended up with a great LPF. GL es 73 Gedas, W8BYA ----- Original Message ----- From: "Edward R Cole" <kl7uw(a)acsalaska.net> To: "Gedas" <w8bya(a)mchsi.com>, moon-net(a)mailman.pe1itr.com Sent: Sunday, August 28, 2016 2:58:35 AM Subject: Re: [Moon-Net] 6m EME Almost QRV Gedas, Interesting info on the transmitting caps. I'm using them in my 2m W6PO-8877 with no apparent issues. The filter I am building is from the QST article in Nov. 2002, pp34-38. This is a LP filter using flat brass plates insulated with teflon sheeting as the capacitors. I will check it on my HP141T using a low level signal. I has a tunable cap across the middle inductor for adjusting 2nd harmonic rejection and SWR at 50-MHz. I will make some photos of the finished filter to place on http://www.kl7uw.com/6m.htm 73, Ed At 04:50 PM 8/27/2016, Gedas wrote: GM Ed, I am glad the information may be of some use and interest for you. You will have to forgive me for sending this to you via the group, my ISP, for some reason will not deliver an e-mail to you directly. I have tried to send you hello's and other RF related mail for several years with zero success. Anyway, What I did to combat the fan noise issue was to purchase (4) 48 VDC box fans then run them at 24-28 VDC. There is still more than enough air flow to keep the PA cool (even running WSJT modes) and the noise goes down by quite a bit. If I recall, the noise is a non-linear function of RPM and you can gain a lot or "quiet" by running them at a lower voltage. The fans do not seem to mind running at the lower voltage as they are cool as well. Just a thought for you to consider. If you have any questions etc feel free to ask, I will be happy to do what I can. Oh, that image of the LPF assembly.....that was an early prototype based on the W6PQL design during the development stage here at this end....the final filter looked very similar but has some different part values. The picture also does not show the other 1/2 of the brass shielded enclosure and dividing wall. I learned something very, very interesting while designing and building this filter. Since I needed a significant amount of rejection in the stop-band region I found that I had to use a dividing wall within the LPF shielded enclosure to reduce some inner-stage capacitive coupling. That helped get me an additional 5-6 dB I.L. and finally got me my -60 dBc specs met. But the most significant finding I had, at least in my case, was that the values of the capacitors I initially used were not very stable as you used then at higher frequencies. Even some expensive, good quality (new) and high-current RF tramsmitting CD caps changed their values as I went up in frequency. This was NOT a thermal issue either. These caps were mica wrapped around by silver sheet metal as plates and are commonly used for RF transmitters etc. In the end, my values were hand picked to optimize the final Bode plot (which also you have not seen). Below is an example of just 3 of the typical caps that I measured. You can see that in my case, for the caps I had on hand, the simple Silver Mica caps proved to be the most accurate and stable in terms of retaining their value as they were used in the VHF region. Most caps are rated (spec'ed at low frequencies). Here is a quick plot showing the drop in capacitance as the frequency was increased: http://w8bya.com/var/albums/Misc/HAM-1/Different-Caps-1.jpg?m=1394322692 Click on the image to expand it. I will see if I can take some images of the final filter and it's brass shielded enclosure with "dividing wall" if you like. 73, Gedas ----- Original Message ----- From: "Edward R Cole" <kl7uw(a)acsalaska.net> To: "Gedas" <w8bya(a)mchsi.com> Sent: Saturday, August 27, 2016 3:54:02 AM Subject: Re: [Moon-Net] 6m-eme almost QRV Gedas, Thanks for the info. I haven't checked bias on individual transistors, yet. Just quick tests to see what it would do. One good indication is no noticeable heat in the exhaust air so must not be driving them too hard. Hell of a noise from those five fans. Pretty much convincing me to remotely mount the PS and amp on the tower at ground level running 240vac and control wires. If I did that I would only need a run of RG-213 for drive power and my transmission loss would be cut in half. I'm measuring RF power with Bird 1000A element so accuracy is unknown. I have a 2500B element and 500H element which I can make comparison readings. I use the 2500B on 2m measuring my 8877. I'm pretty sure is reads low about 1300 when running 1400w. I've compared it at 500w with a 500C element. I'm being more careful with dc current readings since I can measure dc input power easily. Thanks, again. 73, Ed At 04:30 PM 8/26/2016, you wrote: Hi Ed and congrats on getting your Harris up and running. It looks like you are in the final stretch of the project and it will be fascinating to see how your setup performs. I wanted to share with you some data I collected on my Harris 6m SSPA from a couple years ago when I was first bringing it on line hoping you may find some of the information interesting and maybe even useful. Thanks to a wonderful EME friend who let me borrow his high-power Directional Coupler good from 6m on up to 500 MHz (a Werlatone Model C3908), I was able to get a pretty accurate picture of just how bad my particular SSPA was in terms of spectral purity (or actually lack of hi-hi). I found that my solution required me to design and build my own LPF assembly since 2xF0 and 3xF0 were so high. I am not sure if it was just my particular unit that was so dirty or if they are all like this. I made sure before any measurements that all of the MOSFET's were properly biased (400 mA rings a bell if memory serves me right). I also seem to remember 2xF0 was a whopping -23 dBc while 3xF0 was -47 dBc ! Nasty. I know a lot of fellas assume that any available LPF assembly rated for 1 kW will be adequate but after seeing how bad my unit was I knew it was going to take a bit more effort. Some people even think that only the second harmonic needs to be filtered which may be quite incorrect for their unit. I thought I would share with you a couple images of what the raw spectra of my particular Harris looked like going right into a 50 ohm load, then what the spectra looked like after the home brew LPF assembly. The final image show what the Bode plot of the LPF looks like along with the physical implementation. Best of luck. Gedas, W8BYA Gallery at http://w8bya.com Light travels faster than sound.... This is why some people appear bright until you hear them speak. ï»¿ _______________________________________________ Moon-Net posting and subscription instructions are at http://www.nlsa.com/nets/moon-net-help.html 73, Ed - KL7UW http://www.kl7uw.com "Kits made by KL7UW" Dubus Mag business: dubususa(a)gmail.com _______________________________________________ Moon-Net posting and subscription instructions are at http://www.nlsa.com/nets/moon-net-help.html

6 years, 9 months

Re: [Moon-Net] Moon-net Digest, Drive, bias etc. levels for K2RIW 432 parallel

by Jan Lustrup

I was also running a parallel 4CX250B amplifier on 432MHz (K2RIW design) back in the 1990's and had the same problem. Max 400Watts out with 1KW input and the PA got very hot! I tried everything and never got over 40% efficiency, even with new tubes. I was thing of some kind of circular effect in the plate line, so I cut a long line slot with different lengths at a time, but nothing helped. I gave up on that one in the end. I then built a 144MHz (K2RIW design) push/pull version and that one worked like a dream (600-700Watts out). I still have it sitting on my shelf today. I wonder why the 432MHz version never worked like it should. Jan LA3EQ -----Original Message----- From: moon-net-bounces(a)mailman.pe1itr.com [mailto:moon-net-bounces@mailman.pe1itr.com] On Behalf Of moon-net-request(a)mailman.pe1itr.com Sent: 28. august 2016 08:07 To: moon-net(a)mailman.pe1itr.com Subject: Moon-net Digest, Vol 276, Issue 24 Send Moon-net mailing list submissions to moon-net(a)mailman.pe1itr.com To subscribe or unsubscribe via the World Wide Web, visit http://mailman.pe1itr.com/mailman/listinfo/moon-net or, via email, send a message with subject or body 'help' to moon-net-request(a)mailman.pe1itr.com You can reach the person managing the list at moon-net-owner(a)mailman.pe1itr.com When replying, please edit your Subject line so it is more specific than "Re: Contents of Moon-net digest..." Today's Topics: 1. Sensitivity Improvement - Why not Simply a Longer Sequence in JT65? (Jimmy Blanchette) 2. Re: UPDATE - Drive, bias etc. levels for K2RIW 432 parallel 4CX250B amplifier. (Bob Atkins) 3. Re: UPDATE - Drive, bias etc. levels for K2RIW 432 parallel 4CX250B amplifier. (w8zn54(a)verizon.net) ---------------------------------------------------------------------- Message: 1 Date: Sun, 28 Aug 2016 03:08:43 +0000 (GMT) From: Jimmy Blanchette <ve2bmc(a)me.com> Subject: [Moon-Net] Sensitivity Improvement - Why not Simply a Longer Sequence in JT65? To: moon-net(a)mailman.pe1itr.com Message-ID: <7958e5d8-a6f4-4482-a665-7923c146f6c7(a)me.com> Content-Type: text/plain; charset="us-ascii" All, We know that going from JT65B2 (30-sec sequence) to JT65B (1-min sequence) provides a 3dB sensitivity improvement: http://mailman.pe1itr.com/pipermail/moon-net/2012-September/012540.html I am puzzled as to why versions of JT65 have not been developed with longer sequences (e.g. 2-3-4 min, etc.), which would be a very simple way to improve the sensitivity when more of it is needed (e.g. QSO with low ERP station). In addition to the mathematical fact that a longer sequence improves the signal-to-noise ratio "S/N", considering the fact that the libration effect can cause positive swings of many dB in the EME signals which may last for several seconds, wouldn't be possible to take advantage of this natural phenomenon and by averaging several of these longer sequences, extraordinary sensitivity gain could be obtained with relatively little effort? 73, Jimmy KK6FAH

6 years, 9 months

K2RIW Amplifiers

by Byron Tatum

Hello All- Regarding the parallel 4CX250B tube design the best tube in that series (identical physical size) is the Eimac Y-400. I bought some of them surplus years back; they are in silver (paper) or green (metal) colored pop-top cans. At the time I bought mine Eimac could not furnish data as the tube was still proprietary to the USAF, however I was told to consider the tube as a 4CX250R on steroids. I was also told that the tube is rated for 300 mA. class A cathode current. I have used them in VHF/UHF amps over the years and prefer them to the 7580W/4CX250R. Byron W5FH

6 years, 9 months

Drive, bias etc. levels for K2RIW 432 parallel 4CX250B amplifier.

by Bob Atkins

Is anyone here still running a parallel 4CX25B amplifier on 432 (K2RIW design)? I finally got a HV supply that will deliver 2000v at 500mA and I'm seeing about 450W out for around 20W in with around -70v on the grid. The efficiency is around 50%. From published specs I've been able to dig out, it should be capable of 70% efficiency and more power out, but that may require biasing it to -90v and at that my output drops (not enough drive I assume). I'd be very interested in what others are using for bias levels and drive power (and what plate voltage and current). The spec sheet on the 4CX250B says 2000v and 250mA max (per tube), but looking at the old ARCOS data (http://www.k1byd.com/arcosamp.pdf) they have a table of voltages and currents that shows 530W out with 1750v and 640mA of plate currents (key down for 5 minutes). That was with 20W drive and -73v on the grid. 640mA is well over the spec sheet 500mA limit. The original K2RIW article shows 700W out at 70% efficiency with 20W drive and 2000v on the plate and -90 on the grid. That implies a plate current of 500mA I'd be very interested in what others are using for bias levels and drive power (and what plate voltage and current). The tubes are pretty old. This amp was last used (by me) for 432 EME around 1980 and the tubes were used when I got them then, so they are at least 35 years old, possibly 50 years old. I really can't remember what the voltages and power levels were back then. I don't know if these tubes tend to age and if so what the effect of that is. I'd be very interested in what others are using for bias levels and drive power (and what plate voltage and current) and what sort of power and efficiency numbers they are getting. Bob - KA1GT - Bar Harbor, Maine

6 years, 9 months

BF188XR PCB

by Draskovits Gábor HA1YA

Hi, Can somebody say good source for BLF188XR ( or similar) PCB to 6m and 2m amplifiers? BLF188XR ( and similar ) also interested in. 73, Gabi --- Ezt az e-mailt az Avast víruskereső szoftver átvizsgálta. https://www.avast.com/antivirus

6 years, 9 months

Re: [Moon-Net] UPDATE - Drive, bias etc. levels for K2RIW 432 parallel 4CX250...

by TexasRF＠aol.com

Bob, re: tired old tubes. When you measure 500 mA plate current, that is the average value for two tubes. What you can't see is the peak cathode current which is the sum of peak plate and screen current. The plate current peaks can reach about 3X average (even more in class C) and peak screen current peaks can run 5X to 7X average. The cathode has to have enough emission to supply these peak currents. These numbers can be seen on the tube curves where the grid voltage crosses zero and grid current just begins. As a tube wears out, the cathode emission degrades; perhaps that has happened in your case. If you are brave enough, you might test emission by measuring the plate and screen currents while briefly shorting the grid bias voltage to ground. The plate and screen meters will need much higher current ratings and any fuses present would need to be bypassed. This is sort of how tube curve tracers work except they use a pulsed grid voltage source to keep the duty cycle to a low value. 73, Gerald K5GW In message dated 8/27/2016 10:59:20 P.M. Central Daylight Time, ka1gt(a)hotmail.com writes: First, thanks to everyone who has contacted me or posted on this issue. I've gotten some useful links, references and suggestions but I'm still not quite sure why I'm seeing what I'm seeing. In K2RIW's original article he shows a plot of output power vs. drive under conditions of 2000v on the plate, -90v bias and 300v on the screen. He shows 700W out for 20W in (http://bobatkins.com/radio/images/k2riw-output.jpg) Under the same conditions, 2000v on the plate, -90v bias, 300v screen, 5.7v on the filament, I get approximately ZERO output! If I drop the bias back to 70v and can get around 400W out with close to 1000W in (40% efficiency). That's a long way from 700W out and 70% efficiency. It seems to be tuned up properly. Peaking the input shows low VSWR in at maximum power out. Output VSWR is close to 1:1 (lots of coax and a dunny load). He got around 15db of gain. I get arout 0dB under his conditions and around 13dB with the bias dropped to 70v. Input and output tuning caps are all "mid-range" so I'm not running out of tuning range. I'm trying to understand what the cause of this difference is. When tubes get "tired" what are the symptoms? As I said before these are probably tunes from the 1960s and were used when I got them. I don't want to spend $200-$300 on a pair of new tubes only to find that wasn't the problem! For $150 or so I could get a pair of Russian made tubes from Ukraine, but I don't know how they compare in performance to the Eimac tubes. You can get new Chinese made tubes for $50 (http://www.aliexpress.com/item/The-new-export-oriented-4CX250B-tube-4CX250-… l?spm=2114.40010308.4.17.Rjyc93) but I'm not sure I have a lot of confidence in them. 400W out is enough for EME of course, and I'm getting reasonable WSJT echos, but since it is WSJT it's a 100% duty cycle so a bit more efficiency would be nice so I wouldn't be pushing this so hard. At 400W out I'm dissipating almost 600W, 20% over the "limit" for the 4CX250B. I guess the next step is to try driving it harder with the bias at -90v to push it further into class C and see what happens, but if anyone else has been down this road, their comments would be appreciated. Maybe Dick used a pair of Magic tubes, or he was just a much better amplifier builder than me! Has anyone observed the effect of replacing old surplus tubes with shiny new ones? Bob - KA1GT ____________________________________ From: moon-net-bounces(a)mailman.pe1itr.com <moon-net-bounces(a)mailman.pe1itr.com> on behalf of Bob Atkins <ka1gt(a)hotmail.com> Sent: Saturday, August 27, 2016 2:02 AM To: moon-net(a)mailman.pe1itr.com Subject: [Moon-Net] Drive, bias etc. levels for K2RIW 432 parallel 4CX250B amplifier. Is anyone here still running a parallel 4CX25B amplifier on 432 (K2RIW design)? I finally got a HV supply that will deliver 2000v at 500mA and I'm seeing about 450W out for around 20W in with around -70v on the grid. The efficiency is around 50%. From published specs I've been able to dig out, it should be capable of 70% efficiency and more power out, but that may require biasing it to -90v and at that my output drops (not enough drive I assume). I'd be very interested in what others are using for bias levels and drive power (and what plate voltage and current). The spec sheet on the 4CX250B says 2000v and 250mA max (per tube), but looking at the old ARCOS data (http://www.k1byd.com/arcosamp.pdf) they have a table of voltages and currents that shows 530W out with 1750v and 640mA of plate currents (key down for 5 minutes). That was with 20W drive and -73v on the grid. 640mA is well over the spec sheet 500mA limit. The original K2RIW article shows 700W out at 70% efficiency with 20W drive and 2000v on the plate and -90 on the grid. That implies a plate current of 500mA I'd be very interested in what others are using for bias levels and drive power (and what plate voltage and current). The tubes are pretty old. This amp was last used (by me) for 432 EME around 1980 and the tubes were used when I got them then, so they are at least 35 years old, possibly 50 years old. I really can't remember what the voltages and power levels were back then. I don't know if these tubes tend to age and if so what the effect of that is. I'd be very interested in what others are using for bias levels and drive power (and what plate voltage and current) and what sort of power and efficiency numbers they are getting. Bob - KA1GT - Bar Harbor, Maine _______________________________________________ Moon-Net posting and subscription instructions are at http://www.nlsa.com/nets/moon-net-help.html

6 years, 9 months

Jump to page:

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

Moon-net August 2016