Retro 75 Modulation Frequency Response

[k1swl]

gang-

I need to get outside (on snowshoes, probably) and repair the antenna feedpoint. I'll put a Retro on 40M (I have the parts) and listen to band conditions for a few days.  If folks are good with the knowledge that 40M may be a lost cause at night, then I'd offer that version with a clear conscience.

If the activity on 40M is more concentrated at one frequency, the existing 'A/B' switch on the front panel could be used to switch an attenuator in or out.

73- Dave, K1SWL

[Steve - K4HX]

In my experience (since 1976) the bulk (more the 90%) of the 40 meter AM activity I've heard was during the daylight hours. There has been some night time activity (you can find holes between the BC stations) but it's not the time or place for QRP.

[WA3VJB]

In my experience (since 1976) the bulk (more the 90%) of the 40 meter AM activity I've heard was during the daylight hours. There has been some night time activity (you can find holes between the BC stations) but it's not the time or place for QRP.

Down low on 40M has been good at night.

7160Kc is free of international BC, but stateside signals are still well below the levels seen daytime on 40M, so yeah, running low power would be a challenge even when in the clear.

[k1swl]

I should ask before proceeding further with 40M.  7290 kHz appears to be the classic AM frequency for 40M.  Is further down in the band (say, 7160 or 7175 kHz) now the preferred frequency? I have no preference, and would welcome inputs.   

73- Dave, K1SWL
 

[Steve - K4HX]

7290 and 7295 would be a good start.

[WD5JKO]

Hi gang, I finally got my Retro-75 modifications sorted out and documented. This is still a work in progress, but for now it is mostly at the following link:

Open the PDF document first.

http://pages.prodigy.net/jcandela/Retro75/

Jim
WD5JKO

[kg8lb]

Wow Jim ! That was a lot of work. Thank you for taking the time to present your findings and data .

[W1RKW]

Nice work Jim!

I just received a Hammond 106R transformer.  I'm going to give that a try in place of the stock xfmr.    Foot print is slightly different.  I will have to shoe horned in but not a big deal.  It's a potted transformer. Quite dense too.  It feels beefy for something so small.  Will let you know!!

And it can be used up to 50,000ft!

[w8znx]

w4rkw
de w8znx
i waited
but
man
simply had had to build the radio

waiting to hear result

so does the Hammond transformer work
is it worth it

[kg8lb]

W8ZNX DE KG8LB

   Hello Mac. Good to hear from you. I was eager to hear about the Hammond as well but the little Retro does work pretty well in stock form. I have built three of them so far . The first two were in the SWL Labs enclosures but #3 is going in to a custom brass case, will post photos soon . The case is about done but the logos need to be engraved.  Need to stop by the bookstore and get an eyeball, maybe sked a  Retro to Retro local QSO , could be a three way with Mike, N8SDD as he has Retro #2 up in Shelby township MI.

  So far I have fixed the missing ground and added a PTT relay but otherwise as kitted.

[W1RKW]

guys, haven't tried the Hammond xfmr yet.  Been working on something else but it's in the que.

[kg8lb]

Nothing to do with freq response but Retro 75 #3 here at KG8LB is now in it's new homebrew brass enclosure. Still need to engrave the legends on the front and rear plates.


 When we hear a bit more about the great freq response issues we will delve again into old Retro #1 (In the background)

[k1swl]

Gary-

*Very* nice!

Do you do your own engraving? Enquiring minds want to know. :-)

73- Dave, K1SWL

[kg8lb]

Thanks Dave,

  I do everything but the engraving...for now. I have been looking for an old fashioned pantograph to use as opposed to the CNC type routers.  I am lucky enough to have a very fine engraver about 1/2 mile away as well as Royal Oak Name Tag , a maker of just about any type metal label.
  

[W1RKW]

Gary,
You should call it 'Goldfinger'. 

[w8znx]

Ahoy
  finished Retro works fine
best DX so far KA3PTX 438 miles
even better it was a Retro - Retro contact
had to work at it 3x3 each way

yes am thinking of building another
this time making my own box

have lots of other projects

Gary
very nice looking radio

thanks Dave
keep up the good work
have build DSW 40, SW-80,
the Retro,
and just finished a Rock Mite for 80

much fun

yours truly
mac

[WD5JKO]

I have an update on my efforts to enhance the modulation capability with this transmitter. Since my last post, I have seen many instances of the occasional modulation transformer overload. It all depends on the voice, the microphone, level, and which syllable is spoken. The bottom line is that the modulation transformer is just too undersized if you want to pass any lows down to ~ 200 hz.  If you keep C33 at 100 uf, the modifications I made to the original transformer should be fine since the lows roll off pretty good below 500 hz.

Edit: Since writing this post I discovered that the 56mh choke I had across the mic input had opened up. This unloads the mic, and makes a dynamic have a bass boost. That choke used tiny wire, and had about 400 ohms dc resistance.This bass boost made the T3 saturation symptoms all the more worse, and obvious. I wondered why things had gotten worse.  
[/color]
So here is what I did:

I took a 10 watt line matching transformer that I had, and used it as a modulation transformer. The thing was to match a 25v line to 4, or 8 ohms with 5 primary taps to select the power to the speaker.

So I saw the 4, 8 ohm windings were fairly beefy (24 awg wire), and the winding was on the outside. What the heck! Apart it goes, and I unwind 47 turns from 0 to 4 ohms, and 25 more from 4 to 8 ohms.

Then I remove the tiny primary windings (extra tiny wire, maybe 36 awg) to get just a bare bobbin.

Back to the calculator, I got a spool of 28 awg (too small), but two strands of 28 side by side or twisted is ~ 25 awg. Close enough. I take the original wire (24 awg), and wind 50 turns, and bring out the wire. Then I wind 75 more turns (dual 28 awg twisted with drill), and bring that out. Then I wind 13 more turns, and bring that out.

So this gives me an autotransformer with a common, 4 ohm point (50T), a 25 ohm point (50T + 75T), and a 30 ohm point (50T + 75T + 13T).  The turns ratio from the 4 ohm tap to the 25 ohm is 2.5 : 1, and from the 4 ohm tap to the 30 ohm tap is 2.75:1.

So does it work? Well hell yes! I got lucky. All that saturation stuff is GONE. I still have C33 at 330uf, and with a midband peak at 1.5 Khz, going down I see the -3db point at 200 hz, -6 db at 175 hz, and -10 db at 150 hz. It looks like most of the roll-off is R-C stuff in the speech amplifier stages. The thing modulates clean at 100% down to 200 hz. That is really all the low I need, so good enough.

Edit: Looks like the transformer is saturating at 150 hz. I get a big tilt in a square wave drive going below 500 hz, but a sine wave is clean down to 200 hz. I need to ad more turns on the common to 4 ohm tap, and maintain a 2.5:1 ratio. An update to follow.[/color]

Looking at U6 only (remove U5, inject audio into U5 pin 1 socket), I can modulate a decent looking square wave at 1 Khz.

With voice, I think the 25 ohm tap is enough, and I still see > 130 % modulation peaks, but without the ragged occasional overload pattern.

The transformer mounts behind the case, and I'm using one of those two spare holes to hold 1 ear of the transformer.

Edit: The transformer I used was a Speco Model T2510.

http://www.amazon.com/SPECO-TECHNOLOGIES-T2510-LINE-TRANSFORMER/dp/B001MJI8S4

Looking around the net, maybe this one would work as well:

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=300-040

The Speco was not potted, so I could remove the laminations with a small pocket knife to break them free. Somewhat time consuming, but not too bad. Count the turns used from common to 4 ohms, and increase that by about 10%, and figure the 2.5:1 ratio there where the ratio is (total turns) / (common to 4 ohm tap turns). The reason for a few extra turns is that sometimes you cannot get all the laminations back. I got all back except for 1 'E' piece. I did cross interleave, leaving no gap. So 400 ma of DC current seems to NOT cause core saturation, something I was afraid might happen.


Jim
WD5JKO

[kg8lb]

Jim, Here is a nice Shure xfmr : http://cgi.ebay.com/Shure-A102A-70V-Line-Transformer-4-Power-Taps-/170616119239?pt=BI_Circuit_Breakers_Transformers&hash=item27b9835fc7

  Looks like there are plenty of options among the 70V line transformers out there. Good work and very interesting indeed. Just looking for the balance between work, packaging and the return in performance. I am sure we will be making some mods soon.

  Dave, Hope there is still consideration for the 40M version, the upgrades kit etc ! Your little Retro is a screaming success from what I am seeing.


 Mac, Vy good stuff . Will have to get a sked here.  I now have a pair and a third one I built is up in Shelby Twp. at Mike N8SDD's station. We could pick a time and get on with the Retros. About 15 years ago you , Mike and I had a vy good time on a late night meeting up on 10M.

  BTW, I talked with Roy ,K8VWX a couple weeks ago and as a result the king of high power AM is now building a Retro .
 My bud Sam has taken "Two" and "Watter" and contracted that into "T'Watter". The name has kind of stuck around these parts but I don't think Dave is real keen on that monicker .

[W1RKW]

Installed the Hammond 106R into the Retro.  Overall frequency response is about the same as the stock but it does takes a bit more to saturate the transformer. It certainly can be driven harder unlike the stock mod transformer.  Modulation, without the modulation control connected, increased, that is, positive voice peaks in the 1kHz to 2kHz range stand up much taller.  I think with capacitance increases as noted by Jim and others will indeed make the low end response improve.  

Update:
Increased the cap values at C33, C110 and C112 as Jim suggested very early on.  Huge improvement in low end response.  No stability issues at this time.

[WD5JKO]

Bob,

   Good news hearing of your success.  Getting that out of the way allows you to tweak out the rest of the modulation system. Your approach is probably better than mine so long as the Hammond shakes out like you want it to.

   So the audio AGC might need refinement. If you take Q10 collector, and cut the lead at board level, and then insert a resistor back to the trace, then you soften the AGC response which somewhat hides the asymmetrical AGC action of before. I used 1.5K, but 1K seems Ok too:

http://pages.prodigy.net/jcandela/Retro75/AGC_Mod.JPG

   My circuit is different, but as I let more bass through, the AGC ripple voltage (at Q10 base) would come through and modulate the audio signal...a bad thing. Increasing the AGC holdup capacitance will fix this, but also increase the AGC attack time. I did go up to 10 uf though.

I need to revise my Retro75 Hints & Kinks document. My Retro75 is a joy to operate, and I am sure yours is also.

Good Work!
Jim
WD5JKO

[W1RKW]

Hi Jim,
I'm using the modulator AGC system disconnected, it seems to work best for me. Using a scope connected to the Retro I can watch the negative peaks.  But let me tell you with your mods to the 3 capacitors and the larger mod transformer, what a big difference.  Frequency response aside, the modulation % is a lot better especially in the speech range.  Using an outboard processor takes care of the lack of on board mod control.  But if I go portable simply holding the mic away about 6 inches seems to provide good clean modulation.

I'm inclined to try some heising/auto transformer like you did.  With the new transformer mounted off the board by wires it should be easy to do.

Made some progress on my S-meter/IF-AGC mod.  Still tweaking but should have something soon. 

 I beefed up the heat sinks on the mod amp and RF PA.  While testing I could smell the heat.  I know some guys beefed them up so I did the same thing. 

Next will be to work on RCV freq response.