[Amps] GU-84B Conversion

Tue Apr 12 21:44:01 PDT 2011

On 4/12/2011 11:59 PM, Jeff Blaine wrote:
> Yes, pair.  But given 3500z is so easy to get, reliable and cheap, that seems the best way to go unless a guy is just dedicated to
> go ceramic.
>
> The GI46 is a bigger physical tube but it has the same Pd spec of 350 as the GI7b.  So at a higher cost, I'm not sure there is a
> real advantage.  It should be better but if so, the Pd rating should be higher.  I have seen some fantastic claims of far greater
> power out than the GI7b but real data based on actual specs or disection + extrapolation on these tubes is hard to come by.  If the
> Pd is really 350, there is not much of an advantage of the 46 over the 7b.

When it comes to Pd I think of the Eimac 4CX1500B Vs the Chinese 
FU728F.  Electrically they are *almost* identical. The filament voltage 
is different and the FU728F has a pd rating of 1200 watts while the 
4CX1500B has a Pd of 1500 watts, yet the FU728F will run notably more 
power out than the 4CX1500B.

When the physical size of the GI-46 compared to the GI-7B is taken into 
account I have to wonder where they get the figures.  Both tubes have 
the anode coolers added. Is the rating with or without the cooler?  The 
cooler on the GI-46 is larger and more efficient than the cast cooler on 
the GI-7B.  The 1200 watt cooler on the FU728F is larger than the 1500 
watt cooler on the 4CX1500B, but IIRC the 4CX1500B is limited by the 
cathode emission while the FU728F has  plenty of emission.

Typically the better the cooling the more power you can run IF the tube 
has enough emission.
The GI-41 is a pulse tube rated for up to 15KW out, so it has the 
emission.   GI-7Bs have been used in a number of MLA-2500 conversions, 
due to the unavailability of the 8875 tubes *Apparently* they were 
successful.

Another good possibility is the GS-31B which is rated at 1KW Pd and 
available on e-bay for less than $50 USD each...IF you want to purchase 
tubes off e-bay.  Again it'd be easy to change the cooler to a 
transverse unit with a bit of machining. OTOH the stock coolers are big 
and running at the old legal limit, or in an MLA-2500 should work...I think.

And Jeff...My junkbox isn't quite that nice, nor do I have Carl's 
experience. <:-))
As I said earlier, I agree, the best, easiest, and cheapest conversion 
would still be the 3-500Zs.

73

Roger (K8RI)
> The high voltage of the Ps 2900v idle and 2500 at load is perfect for the GI7b - that’s pretty much that tube's sweet spot in that
> the weak spot is the grid drive - higher voltages like this PS has means for a given output, with less Ig.  And longer life.
>
> My feeling is this is the road to an expensive project if you get away from the 3500z option.  For a guy like Roger or Carl with a
> nice junkbox, it's trivial.  Bit if you start spending money, it's easy to end up with a very expensive but pretty unreliable amp.
> I have one here, as a matter of fact.  :)
>
> 73, Jeff ACØC
> www.ac0c.com
>
> -----Original Message-----
> From: Roger (sub1)
> Sent: Tuesday, April 12, 2011 10:51 PM
> To: Jeff Blaine
> Cc: amps at contesting.com
> Subject: Re: [Amps] GU-84B Conversion
>
> On 4/12/2011 11:09 PM, Jeff Blaine wrote:
>> GI7B?
> GI46 would be another choice with a higher plate dissipation rating.
> I've wanted to build up a pair of those, but I have far too many
> projects already under way and too many things that need fixing.
>
> Both could have the anode cooler replaced with one machined for
> transverse cooling.
> Although I think they are passable and "I think" the GI-46 is better
> than the GI-7B I don't know what the specific IM is for either tube.
>
> 73
>
> Roger (K8RI)
>
>> 73, Jeff ACØC
>> www.ac0c.com
>>
>> -----Original Message-----
>> From: Roger (sub1)
>> Sent: Tuesday, April 12, 2011 10:02 PM
>> To: amps at contesting.com
>> Subject: Re: [Amps] GU-84B Conversion
>>
>> On 4/12/2011 9:37 PM, Carl wrote:
>>> Why?
>>>
>>> No room for 3-500Z's or 4-400's?
>> Needs the room for the HV transformer and screen supply?
>>
>>
>> The GU84B is a Tetrode while the Swan is designed around a pair of
>> Triodes. It Is relatively expensive tube even as NOS, sockets run around
>> $150 +/- 30 to 40 dollars. You are likely to have more invested in the
>> tube and socket than the entire amp cost new.    I like the GU-84B, but
>> you are moving up the ladder quite a ways in tube capacity.  You will
>> not need a tuned input and can go with a swamped grid, BUT you may need
>> to add some compensation to keep the SWR/reactance low across the entire
>> operating range.
>>
>> IIRC the GU-84B has been out of production for about 17 years so you
>> would need at least one spare if not two that would or should be swapped
>> out a couple times a year.  They are typically running around $500 USD
>> +/- at present but are trending upwards although Dr Alex has them for
>> under $400USD at present.
>>
>> I would think the conversion from a straight forward Triode circuit to a
>> much more powerful Tetrode would be a major undertaking, particularly
>> when or if the extra protection circuitry (Tetrode control board) is
>> included.  I like Tetrodes, but don't forget the need for a screen
>> supply that will handle both + and - current flow.  Many Tetrodes run a
>> negative screen current when lightly loaded.  At the legal limit output
>> the GU-84 is certainly lightly loaded and with a screen voltage of 375V
>> can run about 80 ma reverse current.  *Claims* of 4 KW PEP out for the
>> tube are common.  A transformer capable of providing enough input with
>> good regulation would be a real hernia builder.
>> OTOH the duty cycle for SSB (20%) and CW(40%) would *somewhat* negate
>> the need for quite that much iron.  Running a Tetrode at less than half
>> it's rated output is not as simple as doing so with a Triode.
>>
>> My guess for your plate voltage you could expect about 250 to 270 ma
>> idle current with proper bias and screen voltages, or roughly 800 watts
>> of no signal dissipation. (any time the amp is keyed unless you use
>> electronic bias switching -EBS) to keep the power low except when a
>> signal is applied.  Many of the older amps would be hard pressed to run
>> that much power with out the EBS.
>>
>> I have to assume (which on this group is never safe) that you are
>> planning on the legal limit output of 1500 watts SSB and CW.  Digital or
>> AM might take a major change in approach, or reduction in power out.
>>
>> How many KVA is the new transformer rated? Barring that, how much does
>> it weigh?
>>
>> You are also going to need to add plenty of air by pressurizing the
>> input compartment.
>>
>> As a personal opinion and all of us on here have different ones, this
>> seems like a a lot of work and possibly expense for the gain.  OTOH I
>> don't know your intentions and reasons for this project.  Like Carl, I'd
>> probably just plug in a pair of 3-500Zs and adjust the bias if necessary.
>>
>> It's much like a friend asked me the other day why I didn't purchase a
>> crank up tower instead of building one. My reply was, because I just
>> wanted to build one. That and it was a really good excuse to get the
>> lathe and milling machine I want.<:-))
>>
>> So good luck on the project for what ever reason you have chosen to go
>> for it.
>>
>> 73
>>
>> Roger (K8RI)
>>
>>
>>> Carl
>>> KM1H
>>>
>>>
>>> ----- Original Message -----
>>> From: "Bill Smith"<ko4nrbs at yahoo.com>
>>> To:<amps at contesting.com>
>>> Sent: Tuesday, April 12, 2011 8:47 PM
>>> Subject: [Amps] GU-84B Conversion
>>>
>>>
>>>> I'm considering converting my old Swan Mark 1 (two 3-400Z tubes) to use
>>>> one GU-84B.  Some time ago I replaced the panel bandswitch in the Swan
>>>> with a Radio Switch Model 86 six position switch with two wafers.  At that
>>>> time I also replaced the tank coil with one that was made for the National
>>>> NCL-2000 amplifier.  My new bandswitch required me to rearrange the RF
>>>> compartment considerably.
>>>>
>>>> The transformer was replaced with a new one from Heyborer in Michigan.  At
>>>> that time I had them use the next larger diameter wire in the HV and
>>>> filament windings when they wound it.  The caps, etc. were all replaced
>>>> with snap in type on a circuit board.
>>>>
>>>> Ok that's the history of the amp to date.  It has a resting plate voltage
>>>> of 2900vdc and 2500vdc or so under load.
>>>>
>>>> Readings with two 3-400Z tubes:
>>>> Plate Current Readings on Operate HV Setting (SSB):
>>>> 270ma using a Tuning Pulser
>>>> Resting Plate Current is 190ma
>>>> 460ma SSB Peak (Talking in Mic)
>>>>
>>>> Plate Current readings using CW Mode(1500vdc)
>>>> Key down 660ma
>>>> Resting 80ma
>>>> SSB 320ma Peak
>>>>
>>>> My question is will the GU-84B perform satisfactorily with this
>>>> transformer?  I know I'll need another filament transformer and one for
>>>> grid voltage.
>>>>
>>>>
>>>> 73,
>>>> Bill Smith KO4NR
>>>> _______________________________________________
>>>> Amps mailing list
>>>> Amps at contesting.com
>>>> http://lists.contesting.com/mailman/listinfo/amps
>>> _______________________________________________
>>> Amps mailing list
>>> Amps at contesting.com
>>> http://lists.contesting.com/mailman/listinfo/amps
>>>
>> _______________________________________________
>> Amps mailing list
>> Amps at contesting.com
>> http://lists.contesting.com/mailman/listinfo/amps
>>
>>
>