Topics

Modamps


Andy G4JNT
 

Over the years I've watched Silicon modamps (the MAR series) morph into the GaAs ERA types, and seen a host of lower frequency very high gain and low noise figure types like the PGA and MGA  ones appear.   All influenced, no doubt, by the mobile market with its low GHz operations

But has much happened at the high frequency end, 8 to 12GHz? What more modern devices are suited up there?

This came about as I wanted a modest gain amp of about 7dB at 10GHz to boost the output of the mixer described in the latest Scatterpoint up to around -4dBm to drive a 1 Watt PA.   I had a couple of ERA-1 devices which, although only specified to 8GHz, have been reported as going to 10Ghz and even higher.

I used a bit of hacked-off old Amstrad board (remember those ?) placing the ERA where one of the old GaAs FETs would have gone, stripping off tuning stubs and adding a couple of 3.3pF ATC capacitors.   Ending up with the arrangement here :   http://www.g4jnt.com/DropF/era1_amp.JPG 

Result,  8dB of small signal gain at 10GHz, which is rather more than extrapolating the data sheet curve suggests. So plenty of margin in hand for my purpose.

But, what modamp type devices are recommended up here?  Surely the ERAs are a bit long-in-the-tooth now ?

Andy  G4JNT


geoffrey pike
 

Hi Andy and group,
I used ERA-1 in a cascaded amp for 3cm and from my notes got 16dB gain but with limited output.
I then used the RFMD NLB-310 device that Farnell use to do.
P1dB +13 dBm and 9.6 dB gain.
I got mine from Australia i think RF-Kits?, not sure if they are still available and RFMD has merged i believe.
The ERA-1 devices are very cost effective and some rudimentary tabs on the lines helped somewhat.
I think there is a picture of the ERA-1 amp in the files section and some other stuff.
regards
Geoff
GI0GDP



On Sunday, 12 June 2016, 10:34, "Andy Talbot andy.g4jnt@... [ukmicrowaves]" wrote:


 
Over the years I've watched Silicon modamps (the MAR series) morph into the GaAs ERA types, and seen a host of lower frequency very high gain and low noise figure types like the PGA and MGA  ones appear.   All influenced, no doubt, by the mobile market with its low GHz operations

But has much happened at the high frequency end, 8 to 12GHz? What more modern devices are suited up there?

This came about as I wanted a modest gain amp of about 7dB at 10GHz to boost the output of the mixer described in the latest Scatterpoint up to around -4dBm to drive a 1 Watt PA.   I had a couple of ERA-1 devices which, although only specified to 8GHz, have been reported as going to 10Ghz and even higher.

I used a bit of hacked-off old Amstrad board (remember those ?) placing the ERA where one of the old GaAs FETs would have gone, stripping off tuning stubs and adding a couple of 3.3pF ATC capacitors.   Ending up with the arrangement here :   http://www.g4jnt.com/DropF/era1_amp.JPG 

Result,  8dB of small signal gain at 10GHz, which is rather more than extrapolating the data sheet curve suggests. So plenty of margin in hand for my purpose.

But, what modamp type devices are recommended up here?  Surely the ERAs are a bit long-in-the-tooth now ?

Andy  G4JNT



Andy G4JNT
 

Found this one at Farnell    http://www.farnell.com/datasheets/77793.pdf

Fixed 5V bias, over 13dB at 10GHz  and a not outrageous noise figure by the look of it.

More expensive than the  ERAs used to be though, but at that gain probably replaces two cascaded..

'jnt


On 12 June 2016 at 10:34, Andy Talbot <andy.g4jnt@...> wrote:
Over the years I've watched Silicon modamps (the MAR series) morph into the GaAs ERA types, and seen a host of lower frequency very high gain and low noise figure types like the PGA and MGA  ones appear.   All influenced, no doubt, by the mobile market with its low GHz operations

But has much happened at the high frequency end, 8 to 12GHz? What more modern devices are suited up there?

This came about as I wanted a modest gain amp of about 7dB at 10GHz to boost the output of the mixer described in the latest Scatterpoint up to around -4dBm to drive a 1 Watt PA.   I had a couple of ERA-1 devices which, although only specified to 8GHz, have been reported as going to 10Ghz and even higher.

I used a bit of hacked-off old Amstrad board (remember those ?) placing the ERA where one of the old GaAs FETs would have gone, stripping off tuning stubs and adding a couple of 3.3pF ATC capacitors.   Ending up with the arrangement here :   http://www.g4jnt.com/DropF/era1_amp.JPG 

Result,  8dB of small signal gain at 10GHz, which is rather more than extrapolating the data sheet curve suggests. So plenty of margin in hand for my purpose.

But, what modamp type devices are recommended up here?  Surely the ERAs are a bit long-in-the-tooth now ?

Andy  G4JNT


Neil Whiting <neil@...>
 

Looks like a useful device, though the 1x0.5mm package might be interesting to handle :-)
Seems to be a 3-lead 0402, so not quite as challenging as some of the other tiny
packages - I'd be prepared to give it a go anyway.

Thanks/73
Neil  G4BRK

On 12/06/2016 10:54, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
Found this one at Farnell    http://www.farnell.com/datasheets/77793.pdf

Fixed 5V bias, over 13dB at 10GHz  and a not outrageous noise figure by the look of it.

More expensive than the  ERAs used to be though, but at that gain probably replaces two cascaded..

'jnt




geoffrey pike
 

This is certainly much easier to work with!
Geoff
GI0GDP



On Sunday, 12 June 2016, 13:45, "Neil Whiting neil@... [ukmicrowaves]"


 
Looks like a useful device, though the 1x0.5mm package might be interesting to handle :-)
Seems to be a 3-lead 0402, so not quite as challenging as some of the other tiny
packages - I'd be prepared to give it a go anyway.

Thanks/73
Neil  G4BRK

On 12/06/2016 10:54, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
Found this one at Farnell    http://www.farnell.com/datasheets/77793.pdf

Fixed 5V bias, over 13dB at 10GHz  and a not outrageous noise figure by the look of it.

More expensive than the  ERAs used to be though, but at that gain probably replaces two cascaded..

'jnt






Andy G4JNT
 

Where's your spirit of adventure !
A 1mm long package with three pads is no big deal ;-)

And nearly twice the gain.

'jnt

On 12 June 2016 at 14:10, geoffrey pike gi0gdp@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

This is certainly much easier to work with!
Geoff
GI0GDP



On Sunday, 12 June 2016, 13:45, "Neil Whiting neil@... [ukmicrowaves]" <ukmicrowaves@...> wrote:


 
Looks like a useful device, though the 1x0.5mm package might be interesting to handle :-)
Seems to be a 3-lead 0402, so not quite as challenging as some of the other tiny
packages - I'd be prepared to give it a go anyway.

Thanks/73
Neil  G4BRK

On 12/06/2016 10:54, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
Found this one at Farnell    http://www.farnell.com/datasheets/77793.pdf

Fixed 5V bias, over 13dB at 10GHz  and a not outrageous noise figure by the look of it.

More expensive than the  ERAs used to be though, but at that gain probably replaces two cascaded..

'jnt







GORDONJ REASON <gordonj.reason@...>
 

Get one of these ...... Excellent value  ........... make's a tough job easier ..........

and get the 0.5 reducer ........... about £20 .... 

The ring light psu , is the best rf wideband source ever ............... so , swopped the dc socket wiring to a positive centre pole ,

and use a "normal" quiet dc supply ............... 


Incidently , for these very small packages , is there perhaps , a small infra red "iron" type device for soldering them in place ....


On 12 June 2016 at 14:43 "Andy Talbot andy.g4jnt@... [ukmicrowaves]" <ukmicrowaves@...> wrote:

 

 

Where's your spirit of adventure !
A 1mm long package with three pads is no big deal ;-)

And nearly twice the gain.

'jnt

On 12 June 2016 at 14:10, geoffrey pike gi0gdp@... [ukmicrowaves] <ukmicrowaves@...> wrote:
 

 

This is certainly much easier to work with!
Geoff
GI0GDP



On Sunday, 12 June 2016, 13:45, "Neil Whiting neil@... [ukmicrowaves]" <ukmicrowaves@...> wrote:


 
 
Looks like a useful device, though the 1x0.5mm package might be interesting to handle :-)
Seems to be a 3-lead 0402, so not quite as challenging as some of the other tiny
packages - I'd be prepared to give it a go anyway.

Thanks/73
Neil  G4BRK

On 12/06/2016 10:54, Andy Talbot andy.g4jnt@... [ukmicrowaves] wrote:
Found this one at Farnell    http://www.farnell.com/datasheets/77793.pdf

Fixed 5V bias, over 13dB at 10GHz  and a not outrageous noise figure by the look of it.

More expensive than the  ERAs used to be though, but at that gain probably replaces two cascaded..

'jnt



 


 

 

 


 


 


Neil Smith G4DBN
 

re: stereo microscopes and teensy devices:  

I do like the look of those devices, thanks for the heads-up Andy.  Not sure what their longevity of supply is though. Mouser show them as obsolete/discontinued, and even the VMMK-3803-BLKG is marked as obsolete, but RS and Mouser have stock of those at about a fiver each.  2dB NF, 18dB gain at 10GHz, same package.  I can’t see anyone doing an alternative, so maybe time to find a cheap source and stock up on them like I did for the old Modamp devices in 1983 or so.  I am down to my last two Avanteks...

I use a scope like this: https://www.amazon.co.uk/7X-45X-Binocular-Stereo-Microscope-Double/dp/B006QXBP6I  and it has been superb for working on even the teensy DFN packages with 0.5mm spacing.  The working distance is about 100mm, which is a bit too close if you are working on boards which are seam-soldered inside a box. The very long reach arm is a joy to use, leaves plenty of free bench space.  I had to drill some small holes in the conical lens shroud to fit my ultra-cheap LED ringlight, but luckily it is radio-silent.

I have a 0.5mm needle tip on a temp-controlled iron that works for most devices, but I have done a few boards using a 100x100x12mm aluminium slab on top of a single mains cooking hotplate.  I use a fine hypodermic to apply medium-temp solder paste to the pads, then pick-and-place the devices under the microscope with tweezers. 

I have drilled a hole into the ally block and fitted a thermocouple so I can ramp up the heat to just under soldering temp, then use a hot-air reflow blower to do the final melt.  I then move the block on to a cooling plate to ramp the temperature down.  Haven’t killed any devices yet.  If/when I mess up, I use one of those 15 quid soldering tweezer things to remove the offending device. Main use of the hotplate is for soldering LDMOS devices to copper heat spreaders, but for those I use 138C melting point solder paste rather than the higher-temp stuff. 

Neil G4DBN



John Fell
 

Have several of the WDG Avantek devices if needed Neil.Trade for real Yorkshire puddings!
John
G0API


SAM JEWELL
 

The GVA123+ from MCL is just over £5.43 in small quantities and comes in the easy to use SOT89 package. It is rated to 12GHz and has useful gain at 10GHz
WA5VJB's little PGA Amp boards, on Arlon low loss material, will take the 89 package. I have lots of the PCBs at £3 each!
73 de Sam





GORDONJ REASON <gordonj.reason@...>
 

where did you fond the 0.5 mm tip and for which iron   ,   Neil    ?

On 12 June 2016 at 18:47 "Neil Smith G4DBN neil@... [ukmicrowaves]" <ukmicrowaves@...> wrote:

 

re: stereo microscopes and teensy devices:  


I do like the look of those devices, thanks for the heads-up Andy.  Not sure what their longevity of supply is though. Mouser show them as obsolete/discontinued, and even the VMMK-3803-BLKG is marked as obsolete, but RS and Mouser have stock of those at about a fiver each.  2dB NF, 18dB gain at 10GHz, same package.  I can’t see anyone doing an alternative, so maybe time to find a cheap source and stock up on them like I did for the old Modamp devices in 1983 or so.  I am down to my last two Avanteks...

I use a scope like this: https://www.amazon.co.uk/7X-45X-Binocular-Stereo-Microscope-Double/dp/B006QXBP6I  and it has been superb for working on even the teensy DFN packages with 0.5mm spacing.  The working distance is about 100mm, which is a bit too close if you are working on boards which are seam-soldered inside a box. The very long reach arm is a joy to use, leaves plenty of free bench space.  I had to drill some small holes in the conical lens shroud to fit my ultra-cheap LED ringlight, but luckily it is radio-silent.

I have a 0.5mm needle tip on a temp-controlled iron that works for most devices, but I have done a few boards using a 100x100x12mm aluminium slab on top of a single mains cooking hotplate.  I use a fine hypodermic to apply medium-temp solder paste to the pads, then pick-and-place the devices under the microscope with tweezers. 

I have drilled a hole into the ally block and fitted a thermocouple so I can ramp up the heat to just under soldering temp, then use a hot-air reflow blower to do the final melt.  I then move the block on to a cooling plate to ramp the temperature down.  Haven’t killed any devices yet.  If/when I mess up, I use one of those 15 quid soldering tweezer things to remove the offending device. Main use of the hotplate is for soldering LDMOS devices to copper heat spreaders, but for those I use 138C melting point solder paste rather than the higher-temp stuff. 

Neil G4DBN


 


 


 


Neil Smith G4DBN
 

Can’t remember where I got the iron from sorry, it was about two years ago, but the tips it uses are these: https://www.amazon.co.uk/Soldering-Tip-Conicle-Solder-KD-M-1/dp/B007EWKMUA

The end of the tip is rounded and seems to work OK with even the smallest SMD Rs/Cs and gullwing chips etc, and it is OK for leadless chips, although something a bit more angular would be better.  I’ve had one tip get stuck to the element and had to slice it off with a Dremel, and one of the five or so I’ve used so far needed dressing a bit to remove a sharp corner.

It has a digital temp display, but no backlight, and it is a pain to use as a result. I keep leaving it switched on overnight because I can’t tell it is on or off….

There is another make which uses the same tips, it has three preset temps and a motion-detecting sleep mode, which would suit me well, so I might keep this one for seam soldering with a fat tip and get one of these for the fine stuff: https://www.amazon.co.uk/PREMIER75W-Digital-Temperature-Controlled-Soldering/dp/B00T0M9XWK.

Not exactly a Weller, but then it isn’t Weller pricing…
Neil


On 12 Jun 2016, at 23:59, GORDONJ REASON gordonj.reason@... [ukmicrowaves] <ukmicrowaves@...> wrote:


where did you fond the 0.5 mm tip and for which iron   ,   Neil    ?


GORDONJ REASON <gordonj.reason@...>
 

Thankyou Neil ...............

I've ordered some tips for my Antex , which I use withan Variac for lower temp stuff ....................

On 12 June 2016 at 23:29 "Neil Smith G4DBN neil@... [ukmicrowaves]" <ukmicrowaves@...> wrote:

 

Can’t remember where I got the iron from sorry, it was about two years ago, but the tips it uses are these: https://www.amazon.co.uk/Soldering-Tip-Conicle-Solder-KD-M-1/dp/B007EWKMUA


The end of the tip is rounded and seems to work OK with even the smallest SMD Rs/Cs and gullwing chips etc, and it is OK for leadless chips, although something a bit more angular would be better.  I’ve had one tip get stuck to the element and had to slice it off with a Dremel, and one of the five or so I’ve used so far needed dressing a bit to remove a sharp corner.

It has a digital temp display, but no backlight, and it is a pain to use as a result. I keep leaving it switched on overnight because I can’t tell it is on or off….

There is another make which uses the same tips, it has three preset temps and a motion-detecting sleep mode, which would suit me well, so I might keep this one for seam soldering with a fat tip and get one of these for the fine stuff: https://www.amazon.co.uk/PREMIER75W-Digital-Temperature-Controlled-Soldering/dp/B00T0M9XWK.

Not exactly a Weller, but then it isn’t Weller pricing…
Neil


On 12 Jun 2016, at 23:59, GORDONJ REASON gordonj.reason@... [ukmicrowaves] <ukmicrowaves@...> wrote:

where did you fond the 0.5 mm tip and for which iron   ,   Neil    ?