The AM Forum
October 14, 2024, 10:36:16 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 ... 3 [4] 5   Go Down
  Print  
Author Topic: Using an SB220 on AM  (Read 105493 times)
0 Members and 1 Guest are viewing this topic.
Steve - WB3HUZ
Guest
« Reply #75 on: April 26, 2010, 09:11:22 PM »

I like the look of that Anniversary Series.
Logged
K5DBX
Guest
« Reply #76 on: April 26, 2010, 09:40:52 PM »

Very kind of you Steve...

Old amps should never die, they should be rebuilt and placed back into service....
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2680


Making AM GREAT Again!


« Reply #77 on: April 27, 2010, 01:51:53 AM »


You can get a glimps of this rebuild over at http://www.k5dbx.com  


VERY Beautiful!

I use an SB220 (not stock, either) on AM, and like it.  I've recently started using a bigger amp, but still love the SB220.

I'm interested in a couple things...  What is the drive level required now, and what is the PEP output...

Also, can you point me to the Eimac publication that talks about directly grounding grids?  I'm a FIRM believer in it myself, but have never run across it from Eimac...  I took it from Bill Orr talking about reducing drive requirements by directly grounding them....  To me it was a bonus the amplifier becoming more stable!


Thanks, in advance....  VERY beautiful construction.  Nice amplifier!


--Shane
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3514



« Reply #78 on: April 27, 2010, 08:28:05 AM »

It was Eimac who recommended the Heathkit method. Rumor has it they changed their mind but Ive been unable to find an official release. And Heath used it for 26 years; its pretty hard to argue with success of around 60K amps.

Carl
KM1H
Logged
WD8BIL
Contributing
Member
*
Offline Offline

Posts: 4412


« Reply #79 on: April 27, 2010, 10:23:47 AM »

Quote
I'd like to see the testing parameters for that. (then) I can't get the IMD to raise any amount measurable with locals listening a few Kc away from me.  No popcorn, etc.


Now there's a definitive test procedure!

Ya see Chris, ya did it all wrong. You probably used a spectrum analyzer and other lab equipment.... dummy!
Logged
ka3zlr
Guest
« Reply #80 on: April 27, 2010, 11:19:36 AM »

There's 2900 members on the Forum now....wholly smokes...

Way Kool Smiley

73

Jack.

Logged
w1vtp
Member

Offline Offline

Posts: 2632



« Reply #81 on: April 27, 2010, 02:43:03 PM »

I wish to He!! I hadn't sold my SB 220 to someone who I considered a "friend."  I gave him a good price 'cause he expressed an interest in it during my last move.  He then proceeded to sell it to someone at a higher cost and my precious SB 220 is gone forever.  NEVER AGAIN!

That P&M said -- I still have my EFJ Courier and it's doing a FB job on AM with my Flex.

Al


* MY FT301 & COURIER.jpg (921.72 KB, 1876x563 - viewed 1139 times.)
Logged
KX5JT
Contributing
Member
*
Offline Offline

Posts: 1948


John-O-Phonic


« Reply #82 on: April 27, 2010, 03:57:32 PM »

Clark says "Tune the amp at low power input on the tune postion. Then when its tuned, go to high power and increase drive.."

I have always heard that this is not the proper way to tune.  I was schooled that the amp should be tuned at max power then loaded a touch more, then reduce the drive to the operating level you want.  Maybe you mean to get a "rough" tune at low power than fine tune at max?  This method is supposed to get the best linear peaks.

John KX5JT
Logged

AMI#1684
K5DBX
Guest
« Reply #83 on: April 27, 2010, 05:41:58 PM »

KM1H, actually it was a the folks at Collins who started to promote the "super-cathode" driven grids. Initially introduced with attempts to tame the 30-L1, engineers at Collins were the ones who prompted those at Heath to use it. Eimac's technical documentation on their design specs for the 3-500z is here.. http://www.umich.edu/~umarc/station/docs/3-500z.pdf

KD6VXI, Shane, thanks for the kind words about the project! When I initially started the rebuild of this project, I began at the drawing board to think about what to do. There is ample documentation and engineering on using low, medium and high Mu triodes. I went back to Eimac’s engineering, and looked at various homebrews in the ARRL archives using various builds using triodes. I know there is a lot of controversy and sides to be drawn in this arena. I worked at Altec-Lansing throughout the 80’s and have years of experience with triodes in the AF range in class A and A/B designs, and know some of the shortcomings. Eimac’s documentation certainly shows that you should firmly ground the grids. When you begin to float the valve, and especially the grids, you have made yourself a "tuned-plate/tuned-grid" oscillator. That’s how you use a low Mu triode designed as an oscillator. That’s why this design can end up making you do all kinds of various voo-doo tricks to keep it from oscillating, when it should have started with the most elegant design, the “grounded grid.”

Working at Altec taught me some very important lessons when it comes to IMD/THD in triode amps. You could take the same parts and give them to the line, and watch the results. More times than not, if the circuit is of a solid design, poor IMD/THD was poor build quality. We know most "splatter" is from either over modulation, or simply running the PS tank circuit out of gas, not from the -40db ~ -50db 3rd and 5th order IMD which is where the 3-500z specs fall. (which is actually better than most transceivers)

To your question of power and drive, it will easily develop 1Kw @ 2.5kv with 100 watts of drive in key down CW, and around 1.35Kw @ 3.1kv at that same level of CW. I have seen on PEP peaks a little over 1.5Kw, but that was without ALC and peaks from the exciter @ 115w out.
 
Logged
Steve - WB3HUZ
Guest
« Reply #84 on: April 27, 2010, 06:46:25 PM »

You are correct John. The plate tuning can be set pretty close to where it needs to be at a lower power, but the load MUST be set at the peak power level you plan to run - at least for best linearity.


Clark says "Tune the amp at low power input on the tune postion. Then when its tuned, go to high power and increase drive.."

I have always heard that this is not the proper way to tune.  I was schooled that the amp should be tuned at max power then loaded a touch more, then reduce the drive to the operating level you want.  Maybe you mean to get a "rough" tune at low power than fine tune at max?  This method is supposed to get the best linear peaks.

John KX5JT
Logged
Jim, W5JO
Member

Offline Offline

Posts: 2508


« Reply #85 on: April 27, 2010, 07:39:16 PM »

I was schooled that the amp should be tuned at max power then loaded a touch more,
John KX5JT

Not to belabor a point, but where did this originate?  I have been fooling with amplifiers for years and never have read it anywhere.  I have a series of technical articles from Eimac and have never read it there either.  What gives, can someone tell me who proposed this, does it work and the theory behind it with proper sources.  What effect does adding more loading have on linearity? 

From what I was taught, the tune cap, coil and load cap are for impedance matching and if you add capacitance to the load side you are just changing the matching impedance which does nothing for linearity.
Logged
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8291



WWW
« Reply #86 on: April 27, 2010, 07:40:25 PM »

I wish to He!! I hadn't sold my SB 220 to someone who I considered a "friend."  I gave him a good price 'cause he expressed an interest in it during my last move.  He then proceeded to sell it to someone at a higher cost and my precious SB 220 is gone forever.  NEVER AGAIN!

That P&M said -- I still have my EFJ Courier and it's doing a FB job on AM with my Flex.

Al

That "friend" thing has happened to me also. Guys like that are pretty low. Thing is, they do not see anything wrong with it.

As for SB-220 on AM, I'd keep it much lower like 150W but then I tend to run things as in CCS mode because they last much longer. I never run my NCL-2000 over 100W carrier for any length of time.

Cooling is the key. Almost every ham-type amp at any power level sufers from inadequate cooling. It's driven by people demanding quiet operation.

That said, enough cooling air can make the difference. I did some experiments with the NCL-2000 at 200W carrier where I set the amp on the top of a bud rack that had a 300CFM rackmount dual squirrel cage blower inside. The amp fit well onto the top opening with the rack's access door opened and although it was noisy, everything in the amp ran only slightly warm to the touch for >1hour including the power transformer and the finals.

At least one person told me that too much air will just blow past the object to be cooled and not pick up any heat, but that was not true in this case. The 8122's might be a special case since the airflow is forced against the convoluted fins by turbulence but that does not explain the transformer's coolness. Maybe the wire was warm, but there is no doubt the core was kept very cool and that has to count for a lot. the NCL-2000 design accidentally lends itself to this technique having many perforated holes and for whatever reason lots of airflow space around the transformer.

I got to think that if the SB-220 (a hackable/non-collectible specimen?) were treated to some mods where 300CFM was forced through it from underneath, it could run reliably at high power for as long as you wished. It would be interesting to get a same-size cabinet, install a bower in it, put it under the SB-220, and push 300CFM up through the amp.

To me the most objectionable blower noise comes from the motor, then the intake, then after that, any "hissing" sound due to restrictions in the airflow.
Logged

Radio Candelstein
K5DBX
Guest
« Reply #87 on: April 27, 2010, 08:22:36 PM »

OpCom, I agree with you concerning the SB220 cooling design. This was one area I chose to make mods that were fairly easy, yet overlooked by many. By replacing the fan blade with the highest CFM pitch design I could find, and shrouding it, made a tremendous difference. Airflow was much more directive, and the chassis design lent itself, once this was done, to draw considerably more air through the transformer side. Noise levels really didn't increase that much either. Certainly a squirrel cage design could deliver more airflow, it would lack the ability to draw air from the right side of the cabinet around the transformer/capacitor area.

Now, where was that plasma cutter I left over here next to the cabinet?! Shocked
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3514



« Reply #88 on: April 27, 2010, 09:17:27 PM »

Quote
KM1H, actually it was a the folks at Collins who started to promote the "super-cathode" driven grids. Initially introduced with attempts to tame the 30-L1, engineers at Collins were the ones who prompted those at Heath to use it. Eimac's technical documentation on their design specs for the 3-500z is here.. http://www.umich.edu/~umarc/station/docs/3-500z.pdf


Actually Collins had no involvement in the 3-500Z or the SB-220, that came directly from Eimac and the chokes were used by several amp manufacturers.  The Collins 30L1 had an obvious influence in the SB-200; different amp several years apart.

That 1980 3-500Z spec sheet has little in common with the original 1968 release which had several different specs and was quite a different tube. That resulted in the 1969 Eimac ads promoting the 3-500Z in the SB-220, Henry, and Swan amps.

Carl
KM1H
Logged
Steve - WB3HUZ
Guest
« Reply #89 on: April 27, 2010, 09:40:16 PM »

Been around since the 20's or 30's. This was standard tuning procedure for linears and grid-modulated rigs.


I was schooled that the amp should be tuned at max power then loaded a touch more,
John KX5JT

Not to belabor a point, but where did this originate?  I have been fooling with amplifiers for years and never have read it anywhere.  I have a series of technical articles from Eimac and have never read it there either.  What gives, can someone tell me who proposed this, does it work and the theory behind it with proper sources.  What effect does adding more loading have on linearity? 

From what I was taught, the tune cap, coil and load cap are for impedance matching and if you add capacitance to the load side you are just changing the matching impedance which does nothing for linearity.
Logged
K5DBX
Guest
« Reply #90 on: April 27, 2010, 09:42:37 PM »

Carl, I think you may be confusing what is being said here. The Collins 30-S1 (which is a tetrode amp) used a floating grid, which is a proper application of this design, since there was zero grid control current. This idea carried over to the 30-L1, and was copied by Heath, as well as other manufacturers. I never said Collins designed the 3-500z or the SB220. I said the idea to float the grids came from Collins engineering. There was an Eimac engineer some years later (late 70's early 80's who was writing for some publications) who was promoting this design concept from the 30-L1, however, it is a bad idea. From an engineering viewpoint, why would you place a triode in a configuration of an oscillator when you don't want it to oscillate? I prefer mine "grounded grid." You can roll your own any way you want. It doesn't matter whether you look at Eimac's 1968 (graphite) or the 1980 3-500z triode specs, "grounded grid" has always meant "grounded grid."

From the 1968 specs..."Operation with zero grid bias simplifies associated circuitry by eliminating the bias supply. In addition, "grounded-grid" operation is attractive since power gain as high as twenty times can be obtained with the 3-500Z in a cathode-driven circuit." (which is what the SB-220 is) In the 1980 revision, they include a "Typical Cathode Driven (grounded-grid) Amplfier Circuit for use with 2 3-500Z Tubes." Please note grids are grounded, not floated.

Logged
Pete, WA2CWA
Moderator
Member

Offline Offline

Posts: 8133


CQ CQ CONTEST


WWW
« Reply #91 on: April 27, 2010, 10:50:56 PM »

Been around since the 20's or 30's. This was standard tuning procedure for linears and grid-modulated rigs.


I was schooled that the amp should be tuned at max power then loaded a touch more,
John KX5JT

Not to belabor a point, but where did this originate?  I have been fooling with amplifiers for years and never have read it anywhere.  I have a series of technical articles from Eimac and have never read it there either.  What gives, can someone tell me who proposed this, does it work and the theory behind it with proper sources.  What effect does adding more loading have on linearity? 

From what I was taught, the tune cap, coil and load cap are for impedance matching and if you add capacitance to the load side you are just changing the matching impedance which does nothing for linearity.

I don't see any modern amplifier manufacturers adhering to this "good old days" procedure. Flipping through about 6 different amplifier manuals, most just say, set drive level of the exciter to some "X' value of initial plate current or some initial grid current, and then quickly adjust Tune and Load or Plate and Load controls for maximum RF output using either an on-board RF output indicator or an inline RF output indicator. None mention any last touch up of the Load control. As Jim points out, I don't see where changing the matching impedance with the Load control, beyond maximum RF output, does anything for linearity. Maybe it's a function of the type of RF output network you have between the final tube(s) and the antenna.
Logged

Pete, WA2CWA - "A Cluttered Desk is a Sign of Genius"
KX5JT
Contributing
Member
*
Offline Offline

Posts: 1948


John-O-Phonic


« Reply #92 on: April 27, 2010, 11:14:25 PM »

Actually the main point in my comment was to tune with max power.  The adding just a touch more on the loading comes from what I've heard from others.  But the using as much drive as the amp takes as a max when "dipping the plate current" is supposed to help to give the best linear peaks. 

Where did I hear this?  I heard this from other hams on the air.  In fact after hearing it I took notice that if I use my rice box (TS-570) at 100 watts CW to tune the Henry 2K4, then switch over to my DX-60 (which in CW will only put out 60 watts) I do typically see better peaks, about 1KW instead of 800W from the DX-60 on AM.  If I use the DX-60 in CW position I get about 800W and a more flat topped peak on AM.

So it does seem to work i my case.  I add just a scoatch more loading because I've heard doing that helps.  I can't really say I see that on the scope, but I only add enough to sacrifice maybe 10 watts peak.

John KX5JT
Logged

AMI#1684
Ott
Member

Offline Offline

Posts: 176



« Reply #93 on: April 28, 2010, 12:33:52 AM »

Been around since the 20's or 30's. This was standard tuning procedure for linears and grid-modulated rigs.


I was schooled that the amp should be tuned at max power then loaded a touch more,
John KX5JT

Not to belabor a point, but where did this originate?  I have been fooling with amplifiers for years and never have read it anywhere.  I have a series of technical articles from Eimac and have never read it there either.  What gives, can someone tell me who proposed this, does it work and the theory behind it with proper sources.  What effect does adding more loading have on linearity? 

From what I was taught, the tune cap, coil and load cap are for impedance matching and if you add capacitance to the load side you are just changing the matching impedance which does nothing for linearity.

I don't see any modern amplifier manufacturers adhering to this "good old days" procedure. Flipping through about 6 different amplifier manuals, most just say, set drive level of the exciter to some "X' value of initial plate current or some initial grid current, and then quickly adjust Tune and Load or Plate and Load controls for maximum RF output using either an on-board RF output indicator or an inline RF output indicator. None mention any last touch up of the Load control. As Jim points out, I don't see where changing the matching impedance with the Load control, beyond maximum RF output, does anything for linearity. Maybe it's a function of the type of RF output network you have between the final tube(s) and the antenna.

Evening Pete... found this on page 13 of a TL-922 operating manual...

Ott



* tl-922.JPG (186.89 KB, 1182x899 - viewed 1154 times.)
Logged
K5UJ
Contributing
Member
*
Offline Offline

Posts: 2814



WWW
« Reply #94 on: April 28, 2010, 01:02:56 AM »

I'm no wizard by any stretch so I'll chime in  Grin  First let us establish a few things.  We're talking about grounded grid, triode, cathode driven linear amplifiers right?  "Increase the loading":  We all understand we're talking about reducing the value of the load capacitor, correct?  There is confusion over this (in some parts; not here I expect) because a lot of amps have the load cap front panel numbers going  up clockwise, and internally a clockwise rotation of the cap shaft reduces the load cap's capacitance.  But, other amps have it right (in my opinion); you turn the big knob to a higher number and the capacitance goes up.  But not everyone knows or understands this, so "increase the loading" comes from "turn the knob to a higher number" but the guy with the oddball amp where that increases the capacitance is thinking, "huh?"   Or you can sort of think of it as increasing the participation of the actual load i.e. the feedline and antenna, doing more with the pi network inductor's stored energy and tangoing less with the load capacitor.    Okay, anyway now that we are past that, why reduce the load cap value?
It increases the average plate current, i.e. the output tank circuit is slightly off resonance but grid current is lowered (which is good) and (I'm pretty dim on this part) internal output tank circuit voltages are lowered slightly (this is good) but more important, with less loading, on peak envelope crests, you have reduced the possibility of breaking into non-linearity.

I'm still trying to figure some of this stuff out so please direct questions to a real expert hi hi  (many are on this board).

Rob
Logged

"Not taking crap or giving it is a pretty good lifestyle."--Frank
WD8BIL
Contributing
Member
*
Offline Offline

Posts: 4412


« Reply #95 on: April 28, 2010, 08:09:04 AM »

Quote
Not to belabor a point, but where did this originate?  I have been fooling with amplifiers for years and never have read it anywhere. 

You can see it first hand at your station with a good scope.
Tune up with a tone (I use 1kHz) at 100% in the "low power" position. Then go to High power and you'll find the load needs tweeked. WHY?

Most hi/lo switched on amps cause an increase in plate voltage. You've just changed the plate impedence!

When you modulate you change the voltage and current. Peak values present a different impedence than nominal.

So yes, it's all about impedence.
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3514



« Reply #96 on: April 28, 2010, 09:46:13 AM »

Quote
Carl, I think you may be confusing what is being said here.


Im not confused and Im fully aware of the timelines of Collins equipment. However the Super Cathode design concept preceeded the 30S1 with Bill Orr and others publishing about it.

The 30L1 grid bypasses came about in an attempt to tame four 30's design 811A's where it worked to some degree in a zero bias amp. In the SB-200 the idea was borrowed for tubes that required an operating and cut-off bias. While it works the amp is not fully neutralized on 15M and above.

Most engineers and some hams are fully aware of the SB-220 and other amps controversy over grid chokes that has been going on for decades. Two of the main protagonists have been at it for decades in forums and web pages. However the chokes in the SB-220 are not the source of instability as they simply act as 25 Ohm resistors at 80-10M and you are left with a resistive loaded capacitive divider that reduces tube gain a bit and promotes stability. The problem is that Heath initially copied the 200pf from the SB-200 and it took until late SB-221 and then the HL-2200 before they realized the error and reduced the values.  The true primary causes of SB-220 series instability is out of tolerance 47 Ohm suppressor resistors and the VHF choke, RFC-2 which just happens to be resonant in the range of the 3-500 parasitic. Replace the parasitic resistors and use a nichrome 10-20 Ohm real wire wound resistor (which does double duty as a HV surge current limiter) in place of RFC-2. No voodo science involved. The whole process can be observed with a GDO and spectrum analyzer and then the cure is simple. The oft arcing bandswitch and Tune cap is a tank circuit design flaw, parasitics are not involved.

The 3-500Z amps with directly grounded grids are no more stable than proper floaters. The AL-82 is a prime example of an IED when it lets loose. The AL-80 family is close behind. The B&W/Viewstar PT-2500 series also had IED problems at first.

Eimac reversed themselves in the late 70's simply because the now legal 1500W 160M rule complicated the design with 160-10M amps and they didnt want to get involved. The fact that they warranteed failed 3-500Z's in Heath and other floating grid amps pretty much says it all.

Carl
KM1H
Logged
Pete, WA2CWA
Moderator
Member

Offline Offline

Posts: 8133


CQ CQ CONTEST


WWW
« Reply #97 on: April 28, 2010, 01:27:46 PM »

[Evening Pete... found this on page 13 of a TL-922 operating manual...

Ott


My take on that TL-922 SSB tuning procedure is that, as you tune the amplifier's Plate and Load controls for maximum RF output, the "last" or "final" adjustment should be made only with Plate control. This is different then the earlier poster's description of dip the plate", increase loading, dip the plate, increase loading, until to get to the specified operating plate current recommended by the manufacturer and then as a final adjustment, increase loading just a hair or two.
If you go to the preceding page on the TL-922 manual, you'll should notice that for the CW tuning procedure, they use the dip the plate, increase loading procedure, and then no final "increase the loading just a hair".
Logged

Pete, WA2CWA - "A Cluttered Desk is a Sign of Genius"
w1vtp
Member

Offline Offline

Posts: 2632



« Reply #98 on: April 28, 2010, 01:57:20 PM »

Here's a web site I've visited many times.  Interesting guy.  Here's his version tuning up a linear amp.

http://www.w8ji.com/loading_amplifier.htm

This is the method I agree mostly with.  I like the idea of using a pulsing circuit to minimize overloading stuff tho'.

Drill up to his home page for some drool pictures of antennas and his workshop

Al
Logged
K5UJ
Contributing
Member
*
Offline Offline

Posts: 2814



WWW
« Reply #99 on: April 28, 2010, 02:48:45 PM »

This is really good information here too:  http://www.cpii.com/library.cfm/9

Rob
Logged

"Not taking crap or giving it is a pretty good lifestyle."--Frank
Pages: 1 ... 3 [4] 5   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 0.069 seconds with 18 queries.