ATS-3 from KD1JV

ats-20mI hesitated ordering the original Sprint and the Sprint II when they made their appearance, thinking they were really just “novelty kits” not intended or capable of anything other than pre-arranged QSOs. But the posts to QRP-L started indicating otherwise and I decided to order my own AT Sprint III.

This is one of the most impressive radios I’ve owned or heard about in my almost 27 years of being in the hobby. I’ll use the page to document my experiences with this amazing radio which, after one week, has far exceeded my expectations.

Building the ATS-3

This was my first attempt at building a kit with SMD parts and I learned a bit along the way from the experience and from other builders who posted their tips to QRP-L. If you’ve already built an ATS-3 and have tips you’d like to pass along, please email them to me. I’ll post them here and give you credit.

  • The main requirements are good lighting, magnification, fine-tipped soldering iron, tweezers and some sort of confined building area. I wish I could remember who posted it to QRP-L, but a suggestion to build this kit in a box top (like a Monopoly game box) saved me on several occassions. The idea is to have an area that contains all your parts, the circuit board, etc. All soldering, parts locating & handling and accidental dropping of parts takes place over this surface for obvious reasons! If you drop a part in a more general area, you may very well never see it again – they are that small.
  • ChipQuik is amazing stuff. Hopefully, you won’t need it but if you do, it is magical in its effectiveness.
  • Some system of separating similar parts of different values is necessary for the caps, resistors and inductors. Not having a muffin tin handy, I used round paper coffee filters. All the 10k resistors in this filter, all the 334k in this one, etc. Also in each filter is a small piece of paper on which is written the value of of those components. Once that’s done, I stack the filters, one inside the other – this takes minimal space and I can easily walk away from the table and then come back knowing right where to begin again. All parts of identical value get soldered before moving on to other values.
  • There are several potential points of confusion. Steve points these out in the manual but they bear repeating. R3, R4 and R11 are yellow. So is L3. If you have 3 of a particular type of yellow, those are the resistors. Also, there are 2 orange resistors (R10, R20) and 2 orange inductors (L1, L2). This is another reason to pre-sort the parts before beginning. Sounds obvious, but “mental discombobulation” can occur. Prevent it by pre-sorting and practicing discrimination – yellow resistors over here, yellow inductor over yonder! It also minimizes the “Now where was I?” effect after returning from a break.
  • L2 on the 80-meter board has 24 turns. For all 24 turns to fit on the coil, you must wind them tightly from Turn 1. There is no spare room on the toroid to accomodate anything less than the most compact winding you are capable of.

specanSpectrum Analysis

The following plots were taken with a recently calibrated Hewlett-Packard HP-8563A spectrum analyzer. As with any other kit, there may be (probably is?) some variation from one ATS3 to the next due to differences in construction from one builder to the next.

To see the spectrum photo for any of the 4 bands, simply click on that band’s thumbnail. Values represent decibels below carrier.

3.5 MHz
7.0 MHz
10.1 MHz
14.0 MHz








There’s a Microscope In Your Camera Bag!

Click on thumbnail for larger image

smttThe solder connections on the SMT chips I’ve mounted so far all look good under a lighted magnifier and they measured zero ohms with my Fluke. But…..behold the magnification capabilities of the modern digital camera. Any component can be easily enlarged to many times its real size (the full-size version of this photo is actually only about 1/6th the size of the original jpeg). This allows me to see an extreme close-up of where I might need to go back and touch up the soldering job without having to squint through a magnifier to see an image not nearly as large, steady or resolved.

A loupe typically magnifies 4x or 5x. With a digital camera and 19″ monitor, I’m getting an effective magnification of about 75x. And the beauty of digital is that is doesn’t cost a cent – just a little added time. Give it a try if you have a digicam.

Most digital cameras are capable of taking close-ups like this that 35mm couldn’t touch without macro lenses, bellows, extension tubes, etc. For this photo, I used a 4-year old (ie obsolete) Nikon Coolpix 995 set at ASA 400, lens wide open and hand-held. Lighting was provided by the 100W (QRO – sorry) bulb in the lamp by my sofa – in other words, low tech to the extreme. Click the thumbnail to see a larger version, where the “microscope advantage” becomes clear.

Bandwidth (Measured on 20m only)

Click for larger


LowBand DXing Results from ARRL DX Contest – ATS3, 80m Dipole & 8 AA Batteries:


I had a bit of time from other chores during the ARRL DX Contest and decided to see what DX I might be able to scare up on the low bands with my Sprint. OK, the truth is, I wanted to see if it was possible to “cross the pond” on 80 meters with QRP and the contest provided the ideal opportunity! Thanks to the Big Guns I worked who also had excellent hearing capabilities.

ATS-3 Builder’s Tips & Modifications

The following QRPers were kind enough to allow me to post their experiences building and modifying their ATS-3′s. Thanks guys.

Mike Caughran KL7R:

Friday – Mounted all active devices. Do not mount the 2N7000s like I did.
Also do not mount anything to the bottom of the board till later.
If you do mount stuff on the bottom you will not be able to use the “pressure gimmick” that AL7FS and others have written about to hold your parts in place.

Be sure to use the super fine solder for the tiny pinned devices first as there is not enough solder to finish the kit and you dont want to be using solder the size of a horse’s leg to mount the DDS device…
3.5 hours to mount the active devices.

Saturday – Mounted inductors, resistors and caps.
Remove only one device at a time from their packages.
Don’t do like I did and empty all the resistors of a given value in preparation for soldering only to have two of them stick to your forearm never to be seen again! Get a good pair of tweezers. I was using hemostats and they were just too bulky.Watch out for the color coding dark green looks a lot like black and the mistake I made is that red is almost indistinguishable from orange.
4.5 hours to mount resistors and caps.

2.0 hours to mount leaded components.

1.0 hour to wire up power plug and test transceiver.
I put a diode in series with the power plug but after looking at the schematic, this is not needed as there is a reverse polarity diode in the circuit already.

45 minutes per board to build & test band modules.

Problems I had:
I lost R9 and R14 but found replacements in an old Soundblaster card.
I broke L3 but made a replacement out of a T37-6.
I had C58 installed 90 degrees out so the receiver didn’t work initially -I found this out by attaching a .1 uF cap to a cliplead attached to my antenna and started poking around till I heard RX noise.

My main problem was that the rig would transmit properly with full output power (It is amazing how much power it gets out from a 9v battery!) but after transmitting, the MPU would reset. This took the hardest bit of debugging to figure out. I started with a voltmeter around the MPU and the voltages all looked as advertized with the exception of TEST. I then put a scope on the RESET pin and noticed that it was being strobbed low after a transmit. I still have no idea why that could be happening and have a note in to Steve but for the mean time, I have just removed R22 so that transistor Q6 doesnt strobe RESET low. The only downside to doing this is that I needed to wait a couple of seconds after a power down/power up sequence.

I started trying to fit it into an Altoids can. Sheesh – that’s going to be tough. Maybe I’ll find a nice plastic case.

Be sure to have more solder and magnet wire on hand.

If you hold DAH on boot up, the dit paddle becomes a straight key for tuning up.

Measurements & Troubleshooting Tips de KL7R:

I measured the RF voltages around the ATS3 with an RF probe for those who may be experiencing delayed gratification.

Be sure to fuse your power. I blew a couple of fuses getting these measurements.

Measurements made with an rf probe like the one at the following URL:
and a digital voltmeter or a VTVM.

7040 with 40 meter module installed

Calibration information:
These are the values measured across a 50 ohm load from an FT 817

5 watt setting 15.1 volts
2.5 watt setting 10.7 volts
1.2 watt setting 6.63 volts
.5 watt setting 4.70 volts

Measurements of the ATS3 using the above probe/meter:

40 mhz ref oscillator OUT (pin5) 1.3V
u10 pin 5 XOUT .6V
u10 pin 6 XIN .6V
U6 pin 8 MCLK 1.3V
u6 pin 19 IOUT 142mV
u6 pin 20 IOUTB .4V
u2 pin 6 OSC 292mV
u2 pin 7 OSC 134mV

For the following measurements:

Disconnect power. Remove filter board. Reconnect power.

u2 pin 6 OSC 195mV
u2 pin 7 OSC 200mV

Disconnect power. Insert filter board. Reconnect power.

Attach a 10uF cap in series with a set of headphones or soundblaster speakers to probe for audio signals.

When you press the MENU button or when you power up you should hear strong CW audio at these locations
The headphone jack
U3 pin 7
U10 pin3 (CAUTION: VERY loud audio here)

Weaker audio can be heard on U3 pins 1,2 and 3.
If you touch your finger to pins 2 or 3 you should hear audio hum at the audio jack.
Transmitter measurements:
(12v source into 50 ohm dummyload)
Some of these are tricky to reach:

Antenna out 11.8 volts
Drains of the 2n7000s 6.0 volts
Gates of the 2n7000s 2.0 volts

U7 pin 1 2.0 volts
U7 pin 2 1.2 volts
Useful gadgets:

A cliplead attached to your antenna at one end and to a .1 cap at the other end. You can use it as a signal generator to walk through the receive RF path.

A cliplead attached to your headphones at one end and to a 10 uF cap at the other end to use as an audio detector.

Gil Stacy NN4CW:

Here are some tips that helped me.

An aluminum rectangular pan with sides helps more than just by trapping parts. The sides of the pan gave my hands a steady rest for parts placement and soldering.

probeTrouble shooting probe: Here’s a photo of a sewing needle probe for clipping to my DVM for testing VDC on smt parts. Try using a regular probe to test U6′s leads with its 10 pins per side on a length of .24″ and you’ll get my drift. ;) The sewing needle is soldered to a piece of copperclad board. Got th e idea from Monty’s site where he has a needle tipped rf probe. The needle will easily slightly penetrate into the solder joint past the flux and will stay put and not slip.

Drilling holes in Altoids tins: Use a drill bit just big enough to accomodate the tang of a rat tail file. Use the tang as a reamer to enlarge drill holes the rest of the way.

Stan McIntosh KD4BTH

The common stepped bits are near-miraculous in how clean of a hole they make in Altoids cans. Have a piece of wood firmly in place directly on the opposite side of where you’re going to drill and then fire away. These don’t grab and peel the metal in the same manner as a conventional bit.

Parts Identification de Bruce N1RX

The writing on many SMT ICs can be hard to read. But, you should not be relying on the printing to determine proper orientation. On some runs of ICs, the direction of the printing varies. Instead, there are one or two methods you can use to be sure you’ve “got it right”:

  • Look for an indented “dot” marking the pin 1 corner. Not a printed dot, but an actual recessed indentation. I use a strong light, at an angle, and a hand-held 10x magnifier to be sure I have this. On my ATS-3, thee IC marked this way include: U3, U5, U6, U8, U9 and U10.
  • The second method used on some ICs is to mold the case of the IC such that the edge of the package that runs along the pin 1 side is beveled. Let’s say you have an 8pin IC like a SA612A mixer. This means that along the side of the IC that has pins 1-4, the top edge of the IC body is cut at an angle. There is actually a “slope” along that upper edge that slants down toward the row of pins. Once you find the edge that has this slope, you have the side of the IC that has pin 1. On my ATS-3, ICs made this way include: U1, U2, U3 and U7.

U4 is non-symetrical (2pins on one side, 3 pins on the other) so placement is not a problem.

As you can see from the above list, some ICs use BOTH methods for indicating pin one. Again, a strong, angled light, and a powerful magnifier are a must for this.